Next Article in Journal
Jellyfish and Ctenophores Around Gotland in the Baltic Sea—Local Data Contributing to Global Assessments
Previous Article in Journal
Reducing Wave Overtopping on Rubble Mound Breakwaters Using Floating Kelp Farms
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

A Comprehensive Review of Ship Automatic Rudder Technology: Development, Challenges, and Prospects

Navigation College, Jimei University, Xiamen 361021, China
*
Author to whom correspondence should be addressed.
J. Mar. Sci. Eng. 2025, 13(5), 851; https://doi.org/10.3390/jmse13050851
Submission received: 20 February 2025 / Revised: 14 April 2025 / Accepted: 21 April 2025 / Published: 25 April 2025
(This article belongs to the Section Ocean Engineering)

Abstract

:
Ship automatic rudder systems are critical to enhancing the precision, safety, and intelligence of modern navigation, which plays a significant role in optimizing vessel control, reducing crew workload, and improving operational efficiency. First, this paper undertakes a meticulous analysis of the evolution of automatic rudder technology research through the application of bibliometric analytical methodologies. Second, a thorough literature review is conducted, focusing on the fundamental components of automatic rudder technology, encompassing sensor technologies, control algorithms, and basic functionalities, including both the latest advancements and the distinct traits of various technological approaches and algorithms. Third, the core functions of automatic rudder technology are analyzed, which includes heading control, wave compensation, ship collision avoidance, safety control mechanisms, path planning, and tracking control. Lastly, this paper summarizes the principal challenges encountered during the development of automatic rudder technology and proposes potential directions for future advancements.

1. Introduction

The safety, economy, and intelligence level of ships are becoming more and more important to the shipping industry. The performance of the rudder, which is the primary piece of equipment used to control the ship’s heading during navigation, is closely correlated with the ship’s navigation stability, heading precision, and capacity to handle challenging sea conditions. The main motivating factor now is the quick development of ship intelligence, and automatic rudder technology, which is a crucial demonstration of this intelligence, is the future of ship control [1]. China has implemented a number of ship intelligence regulations and strategic initiatives to keep up with global developments. For instance, intelligent ships are specifically listed as one of the ten major development fields in the “Made in China 2025” strategy, demonstrating the great importance that the state attaches to this field. China Classification Society (CCS) has carried out a detailed grading of intelligent ships, in which automatic rudder technology is included in different levels of intelligent ship development. Specifically, this grading system includes: Level 1: Basic automatic rudder function optimization to attain heading stability and preliminary automation; Level 2: Remote monitoring and supported decision-making to enhance rudder efficiency and navigational safety through data analysis; Level 3: Fully autonomous navigation control, which implies that, in certain situations, the ship can rely on automatic rudder technology to navigate and avoid collisions. These classifications reflect the comprehensive integration of automatic rudder technology and illustrate the continuous progress being made in the development of intelligent ship systems in China.
With the progress of marine technology, it has become more challenging for the conventional automatic rudder control approach to satisfy the demands of contemporary ships for intelligent maneuvering, high precision, and high adaptability. Therefore, a new phase of ship autonomous navigation technology development has begun with the emergence of automatic rudder solutions that incorporate advanced technologies including deep learning, predictive control, fuzzy logic, and reinforcement learning. For instance, the automatic rudder system can forecast the future state of navigation more precisely by incorporating a convolutional neural network (CNN) to analyze environmental data like sea state, wind direction, and current speed in real time, which enables the system to modify the heading and speed beforehand, ensuring navigation stability and safety [2]. To achieve smooth tracking and effective trajectory adjustment, the system can simultaneously optimize the current control strategy based on the prediction of the future state when integrated with the predictive control algorithm [3]. Besides, the autonomous rudder exhibits more robustness and flexibility in handling a varied and complicated navigation environment owing to the use of fuzzy logic control. By fuzzifying the input variables (such as heading deviation, speed change, etc.) and reasoning with fuzzy rules derived from experts’ experience, it can effectively respond to sudden variations in wind and waves or changes in ship dynamic characteristics [4,5]. Additionally, the introduction of reinforcement learning technology enables the automatic rudder systems to self-optimize the control strategy by a continuous trial-and-error approach as well as learning [6]. Through continuous interaction with the environment, the system can accumulate navigation experience, improve its ability to make autonomous decisions and cope with complex situations, and ultimately, make automatic ship navigation more intelligent and efficient.
The ship’s automatic rudder system is gradually becoming more intelligent and autonomous. To detect the ship’s position, speed, and heading in real time, modern ship autonomous rudders use a variety of sensors, including a gyroscope and precise GPS tracking. They also include sophisticated environmental awareness and collision avoidance technologies. The automatic rudder system can use dynamic collision avoidance algorithms to quickly calculate and avoid the risk of collision to ensure navigation safety. It can also monitor other ships, buoys, reefs, and other potential obstacles in the waterway in all weather conditions using the global navigation satellite system, laser radar, rudder angle sensor, and other sensors [7]. In the complex and changing navigational environment, the automatic rudder system shows powerful capability for planning the regional routes. It generates the optimal sailing route to the destination according to the real-time environmental information and the current state of the ship, and flexibly responds to unexpected situations and unpredictable conditions of the sea [8]. To further enhance the intelligent level of the automatic rudder system, modern ships are also equipped with the ability to communicate with the shore control center in real time. This remote monitoring and command system enables the shore control center to grasp the ship’s status in real time and send control orders or adjust the sailing plan as needed, providing a double guarantee for the safe navigation of the ship. In addition, the close collaboration of sensing technology, path tracking algorithm, tugboat assistance, and dynamic positioning system together constitute a powerful technical support system for the ship automatic rudder system.
Figure 1 depicts the primary framework of this article, which aims to provide a thorough discussion of the fundamentals and development trends of ship automatic rudder technology. The first part elaborates on the significance of ship autonomous rudder technology to improve the degree of automation, efficiency, and safety in navigation by using the analysis tool of a large literature data set. The second part methodically sorts out the main technologies, such as control algorithms and sensing technology, that make up ship autonomous ruddering. The third part explores the fundamental operations of the ship autonomous rudder and the main technologies of tracking control, heading control, safety control, wave compensation, and ship collision avoidance. In the fourth part, this paper analyzes the current challenges of ship automatic rudder technology and provides potential research directions and solutions for this technology, aiming to promote the continuous improvement and innovation of automatic rudder technology. Finally, the future development trend of ship automatic rudder technology is analyzed. In conclusion, this paper provides a comprehensive and in-depth perspective for scholars about ship automatic rudder technology.

2. Analysis of Literature Dataset

Applying bibliometric methods to explore the vast field of ship automatic rudder technology makes it possible to grasp the whole picture of ship automatic rudder technology quickly and accurately, so that the developmental lineage, core issues, active research topics, and the direction of technological evolution are depicted quickly. Scientific tools for visual analysis of literature such as CiteSpace 6.3.R1 and VOSviewer 1.6.20. 126 were utilized to analysis. High-quality articles from Web of Science and CNKI that were closely related to automatic rudder technology for ships were carefully selected to ensure the accuracy and representativeness of the analyzed results. Subsequently, these literature data were transformed into intuitive and easy-to-understand visual information in the form of maps of scientific knowledge, through which the process of developing ship automatic rudder technology was organized systematically, facilitating the effective integration and in-depth analysis of complex information among scholars.

2.1. Keywords

An intuitive and informative heat map shown in Figure 2 is generated by importing the selected literature into VOSviewer software for collaborative keyword analysis, which shows the co-occurrence density and association pattern of the keywords in the selected literature. In the center of the heat map, the keyword “ship’s heading” is particularly prominent, which clearly indicates the core research direction in the field of ship automatic rudder. It is worth noting that the density of “guidance rate” is slightly lower than that of “ship’s heading”, but its prominence indicates that the rapid development of ship automation and intelligent technology is leading to more urgent demand for intelligent navigation systems that integrate advanced algorithms and autonomous decision-making capability. Intelligent navigation, as a natural extension and advanced form of automatic rudder control technology, not only inherits the basic functions of the automatic rudder in heading keeping and tracking, but also significantly expands the autonomy and adaptability of the system on this basis. It deeply integrates cutting-edge technologies, such as artificial intelligence, big data processing, machine learning and advanced control theory, realizing accurate perception, dynamic decision-making and intelligent response to the complex marine environment. In the process of profound exploration of the technology of ship automatic rudders, “control input” and “rudder speed”, as two synergistic keywords, affect the precise control of the automatic steering system to the disturbance input and its feedback output, which constitute the cornerstone of the autonomous navigation system to ensure the efficiency and accuracy of navigation. The integration of these two technologies not only reflects the ability of the ship’s automatic rudder to perform difficult operations in complex marine environments, but also signals the trend of ship navigation control to a higher level of intelligence. From the perspective of the evolution of automatic rudder technology, this process profoundly reflects the transformation from the traditional control mode based on structured rules and preprogramming to the modern, dynamically adjusted, intelligent decision-making, and highly adaptive system.

2.2. Trends of Development

The keywords are extracted and analyzed by exploring the literature dataset in the field of marine automatic rudders through CiteSpace 6.3.1. They are ranked according to the salient intensity in the literature set, and transformed into Figure 3, which clearly shows the dynamics and evolutionary trends of the research in the field of marine automatic rudder technology during the period of 2010–2024. The bar chart in Figure 4 shows the main sources of the literature collected. In Figure 3, some of the keywords have risen significantly since 2018 and reached peak research intensity in 2020, which reveals the trend of accelerating the transformation of ship automatic rudder technology towards intelligence and diversification. The research intensity of the keyword “heading control” is 1.59 in 2022, which indicates that in the field of automatic rudder, heading control still maintains a high research intensity in recent years. Intelligent automatic rudder not only represents the latest development direction of ship navigation control technology, it also predicts that future marine navigation will rely more on the systems highly integrating with artificial intelligence. In Figure 5, “controller design” is another major keyword, which maintains high heat throughout the research cycle. During the period from 2013 to 2018, its research intensity increases steadily, and the rise of “control strategy” plays a complementary role. The continuous innovation of control strategy technology has laid a solid foundation for the stable operation of the automatic rudder in the complex and changeable marine environment, which greatly improves the accuracy of the ship’s navigational tasks and ensures the safety and efficiency of the navigation.

2.3. Related Technologies

A dynamic map of the technological evolution in the field of ship automatic rudder is drawn by using visual collaboration technology VOSviewer, which is shown in Figure 6. The colorful nodes symbolize the different technological methods. The connection between them and the color changes reveal the development of these technologies over time and their mutual influence. In recent years, “path planning”, “wave perturbation”, and other keywords in the map signify that the ship’s automatic rudder technology is becoming more intelligent and adaptive. In the core field of automatic rudder technology, “control accuracy” still occupies an important position, and its continuous process of optimization and improvement is the subject of ongoing investigation for navigation performance. Meanwhile, “Model Predictive Control”, with its foresight and accuracy, has become one of the key technologies for realizing efficient and stable navigation. The combination of these control methods ensures that the ship can maintain the best sailing condition under complex marine conditions. It is worth mentioning that the rapid development of the technologies with environmental perception has added important and safe barriers to ships’ automatic rudder systems. They not only enhance the automatic rudder’s ability to sense the surrounding environment, but also enable it to identify and avoid potential hazards in real time, thus ensuring the safety of navigation. In addition, the profound application of “adaptive control” and “fuzzy control” reflects the flexible response capability of the automatic rudder system to the uncertainty of the marine environment. These control strategies can adjust the control parameters and strategies automatically according to the changes in the actual marine conditions, ensuring that the automatic rudder system can maintain stable performance under various complex conditions. The integration of “neural network” has opened a new chapter in the intelligent development of ship automatic rudder technology. Through continuous learning and optimization, the automatic rudder system can gradually improve its own performance and achieve more accurate and efficient navigation control. Finally, the map also shows the application of key technologies such as “navigation”, “ship collision avoidance” and “trajectory tracking” in the ship automatic rudder system, which constitute the core functional framework of the automatic rudder system, ensuring that the ship can realize accurate navigation and automatic berthing according to the predetermined route and trajectory. The continuous development and improvement of these technologies will further enhance the level of intelligence and comprehensive performance of the ship’s automatic rudder system.

3. Automatic Rudder Technologies

This section provides a profound discussion of the core components of ship automatic rudder technology, including sensing technology, control algorithms and other prominent aspects. By analyzing the latest research results and application examples in these fields, this section aims to reveal the latest progress in these fields and point out the research direction for future research, promoting the continuous progress and innovation of ship automatic rudder technology.

3.1. Sensor Technology

In 2023, the China Classification Society (CCS) formally launched the “Guide to Automatic Rudder System for Ships”, which defines the core application and technical standards of automatic rudder in ship navigation. The sensing technology for a ship’s automatic rudder, as the neural junction of ship intelligent navigation, is the core power to drive the automatic rudder system for accurate execution and intelligent decision-making. It integrates a variety of high-precision sensors and advanced data processing technology to build an all-round, multi-level navigation perception network [9]. In the system of modern ship autopilot, sensor technology occupies a pivotal position, and its application not only profoundly affects the navigation and control ability of the system, but also improves the safety and operational efficiency of the ship. These diversified sensors, such as gyroscopes, magnetometers, and other directional sensors, as well as GPS speed sensors and meteorological and hydrological environment sensors, build a comprehensive perception network of ship navigation. The direction sensor captures the heading changes in real time and provides precise direction guidance for the automatic driving. The speed sensor accurately measures the ship’s speed and position, which is the basis for track projection and heading adjustment [10], and the environmental sensor constantly monitors the weather and water conditions, using environmental data to provide valuable support for navigation decisions. To further improve the navigational accuracy and safety of the system, the ship autopilot system adopts sensor data fusion technology. By integrating multi-source sensor data, such as direction, speed, and environmental information, and advanced data fusion algorithms, this technology realizes comprehensive and accurate monitoring of navigation status. This process not only enhances the system’s ability to perceive the complex navigational environment, it also provides strong data support for intelligent decision-making algorithms, which enable the system to respond quickly in the ever-changing marine environment and effectively reduces the risk of collision, thereby safeguarding navigation safety. In addition, sensor technology promotes the adaptive control of ship automatic driving systems. The adaptive control system can flexibly adjust the control strategy based on real-time sensor feedback to cope with the challenges of complex navigational situations. In particular, the application of adaptive fuzzy control algorithms enables the system to dynamically optimize the control parameters based on the consideration of nonlinear factors, such as input saturation and external disturbances, to enhance the stability and response speed of the system [11]. This ability not only enhances the ship’s navigational performance, but also effectively reduces the risk of accidents caused by human error.

3.1.1. Global Navigation Satellite System (GNSS)

Innac et al. [12] proposed the problem of insufficient positioning accuracy and reliability of the traditional single satellite navigation system in urban canyons, dense forests, and other areas where satellite signals are limited, and GNSS was proposed as a broad concept that synthesizes multiple satellite navigation systems around the world. As shown in Figure 7, GNSS includes the GPS system, Russia’s GLONASS, Europe’s Galileo, and China’s BeiDou, etc. This multi-system greatly enhances the global position, effectively compensating for the shortcomings of a single system in the challenging environments. To further improve the positioning accuracy to the level of centimeters and meet the demand for high-precision positioning, Zhang et al. [13] introduced the technique of real-time dynamic positioning called RTK, which significantly improves the positioning accuracy of GNSS by combining it with real-time and correct information provided by the ground base station. This is particularly important for ship positioning and autopilot, as ships may temporarily lose direct line of sight to satellites when traversing specific environments. The synergy between RTK and GNSS enables the ship autopilot system to obtain accurate position information in real time and adjust the heading accordingly, ensuring accurate navigation of the ship in complex environments. In terms of path planning, Zhao et al. [14] pointed out that GNSS data is an indispensable part of the autopilot system. By combining the positioning information from GNSS, the system can intelligently calculate the optimal route and effectively avoid obstacles and other ships, thereby reducing navigation time and fuel consumption. In addition, Zhang et al. [15] pointed out that modern ship autopilot systems have also widely adopted multi-sensor fusion technology, which combines GNSS with data from other sensors to achieve a comprehensive perception of the surrounding environment.

3.1.2. LiDAR

An innovative solution is proposed by Pendleton et al. [7] to deal with the serious challenge of insufficient navigation accuracy for ships in complex waters and crowded environments, which uses the LiDAR systems to map high-resolution 3D environments in real time. A schematic diagram of LiDAR operation is shown in Figure 8. By accurately capturing the return time and reflection angle of laser pulses, this technology provides ships with detailed information about their surroundings, thus significantly improving the navigation accuracy of the autopilot system. This innovation not only effectively reduces the risk of collision, but also greatly enhances navigation safety. Further, to address the problem of insufficient accuracy of heading control for ships in complex and changing navigational environments, Kim et al. [16] explored the potential for profound processing of real-time LiDAR data. They found that by dynamically adjusting the heading in real time using LiDAR data through an autopilot system, the ship could maintain a stable navigational attitude in changing environments. In extreme environments where bad weather and complex ocean conditions are intertwined, the guarantee of navigational accuracy is especially critical. Liang et al. [17] have perceived this challenge and proposed LiDAR as a stable and reliable measurement tool. They pointed out that LiDAR can maintain measurement accuracy in the midst of fierce winds and waves, and ensure that the ship always travels safely along the predetermined route by using real-time monitoring of environmental changes and intelligent adjustment of control parameters. This feature enables ships to respond flexibly and navigate efficiently in different marine environments. To further improve the performance of ship autopilot systems, Wang [18] focused on the optimization of control strategies. They combined LiDAR with advanced control strategies, such as model predictive control (MPC), which significantly improve the response speed and stability of the system by predicting the future state of the ship and making adjustments in advance. This innovation not only solves the problem that the control strategy is not advanced enough to cope with complex ocean dynamics, but also points out the direction for the future development of ship autopilot technology. In addition, to address the limitations of insufficient information from a single sensor, which makes it difficult to perceive the environment comprehensively, Zhang et al. [13] proposed a solution of data fusion technology. They emphasized that by fusing LiDAR data with multiple sensors, such as GPS and inertial navigation systems, more comprehensive and accurate environmental information support can be provided to the autonomous driving system. This cross-sensor data fusion not only improves the system’s level of intelligence, but also equips ships with a stronger ability to respond autonomously in complex environments. Finally, given the global consensus on the need for shipping sustainability, Zhang et al. [19] have thoroughly explored the potential of LiDAR technology in promoting the development of green shipping. By optimizing routes, improving fuel efficiency, and accurately monitoring changes in the marine environment, LiDAR technology can reduce carbon emissions from ships, promote compliance with environmental protection regulations, and contribute to the sustainable development of the shipping industry.

3.1.3. Rudder Angle Sensor

Rudder angle sensors play a crucial role as core components of ship autopilot systems. As shown in Figure 9, a typical rudder angle sensor provides real-time feedback, allowing the autopilot system to rapidly detect changes in rudder position and adjust heading accordingly. This capability is essential for maintaining course stability in complex marine environments. As pointed out by Liu et al. [20], this real-time capability is crucial for enhancing the immediacy and accuracy of navigational control. With the evolution from traditional mechanical and electronic sensors to software-based Virtual Rudder Feedback (VRF) technologies, rudder angle sensing has undergone significant changes in both design philosophy and operational constraints. Mechanical transducers measure rudder displacement physically, while electronic transducers utilize signal conversion to improve accuracy and transmission efficiency. To accommodate diverse vessel configurations, particularly small-sized or unmanned surface vehicles (USVs) with strict spatial and structural limitations, VRF systems simulate rudder angle using software algorithms. This eliminates the need for physical sensors and enhances system flexibility, maintainability, and integration potential [21,22]. Design factors such as mass, inertia, and mission profile must be considered when choosing suitable sensing technologies [23]. Recent research has further explored advanced control strategies that integrate rudder angle sensors into intelligent systems. On the one hand, adaptive fuzzy control and finite-time adaptive control methods have been proposed to improve response time and heading stability, especially under rudder saturation or when heading rate measurements are unavailable [24]. On the other hand, integrated design approaches are gaining attention, aiming to reduce overall system complexity by combining rudder sensing with other navigation sensors, thereby improving reliability and synergy [25].

3.2. Control Algorithms

3.2.1. PID Control

As shown in Figure 10, the algorithm effectively controls the system by calculating the error between the desired output and the actual output and adjusting the control signal according to the proportion, integral, and differential of the error. The PID controller remains one of the most widely used approaches in ship heading control due to its simplicity and robustness. It also serves as the foundation for conventional marine control systems [26]. A PID controller can precisely control the ship’s heading and ensure stable navigation on the set course. By adjusting the rudder angle in real time, the PID controller can respond quickly to external disturbances, such as the influence of wind and water currents. In complex navigational environments, PID controllers can realize path tracking to ensure that the ship follows a predetermined trajectory [27]. PID controllers can be applied to many different types of control systems, including linear and nonlinear systems, which gives them great flexibility in ship autopilot [28]. However, in the face of the inherent nonlinear characteristics of ship motion, variable external disturbances, system inertia, and other complex challenges, the traditional PID controller is often inadequate in practical applications, which can easily lead to overshooting, steady-state error, and so on [29]. To overcome these limitations, scholars have continuously explored and proposed a variety of optimization strategies. On the one hand, hybrid control strategies, such as the deep integration of fuzzy logic and PID, not only retain the simplicity and efficiency of PID control, but also leverage the powerful ability of fuzzy logic to effectively deal with the nonlinear factors in the motion of the ship, which significantly improves the adaptability and robustness of the control system [30]. On the other hand, the introduction of adaptive control technology, especially the adaptive PID controller based on BP neural networks, realizes the accurate capture and fast response to the ship’s sailing state through real-time adjustment of the control parameters, which further improves accuracy and efficiency of the control [31]. In addition, the optimization of PID control algorithms is reflected in several aspects, including the fine-tuning of proportional, integral, and derivative gains, the integration of model predictive control (MPC) strategies for anticipatory adjustment, and the application of data-driven methods for adaptive parameter optimization under varying operating conditions. Modern optimization algorithms, such as particle swarm optimization and ant colony optimization, provide a powerful tool for fast and accurate tuning of PID parameters, which significantly improves the overall performance of the control system [32]. The model predictive control combined with the neural network approach achieves effective responses to uncertainty and delays problems by predicting the future system behavior, which further enhances the intelligence and adaptability of the control system. The PID parameter adjustment strategy based on real-time data analysis also provides data support for the continuous optimization of the control system [33]. In complex marine environments, the motion characteristics of a ship may change with time and conditions, so it is essential to design controllers that can cope with static environments and adapt to dynamic changes. By introducing an adaptive mechanism, PID controllers can maintain good performance under various sea conditions and improve the safety and economy of ship navigation [34]. By continuously integrating modern technology, intelligent control concepts and adaptive mechanisms, scholars are gradually breaking through the limitations of traditional PID controllers, which establishes a new path for the future development of ship autopilot systems.

3.2.2. Fuzzy Control Algorithm

The fuzzy control algorithm, renowned for its robustness and adaptability, exhibits exceptional performance in ship control systems, particularly in handling system nonlinearities and coping with external environmental uncertainties. Robust fuzzy control frameworks for nonlinear marine systems have also been studied in earlier works. Lin [35] introduced an adaptive fuzzy control approach capable of handling unknown dead zones, which remains relevant for modern surface vessel applications. Its schematic diagram is shown in Figure 11. During ship navigation, the fuzzy control algorithm enables real-time monitoring and flexible adjustment of the vessel’s heading, thereby enhancing the overall safety and stability of the navigational process. By continuously optimizing the control strategy, the fuzzy control method effectively mitigates deviation and instability during navigation. While capable of achieving high control accuracy in suitable configurations, its performance is sensitive to the design of the rule base and membership functions. Poorly tuned fuzzy systems may exhibit degraded tracking accuracy, particularly under rapidly changing external disturbances or poorly modeled nonlinearities [36]. To cope with the challenges of nonlinearity and uncertainty prevalent in ship autopilot systems, Yang et al. [37] proposed a model reference adaptive robust fuzzy control strategy. The strategy skillfully combines the fuzzy logic system with the theory of Lyapunov stability, and successfully realizes the accurate approximation of the complex and unknown set of total dynamic characteristics in the system of ship. The experimental results show that the algorithm not only firmly ensures the asymptotic stability of the ship steering autopilot system, but also demonstrates the excellent performance of near-zero tracking error, which greatly improves the accuracy and safety of ship navigation.
In recent years, with the continuous development of ship autopilot technology, scholars have proposed many adaptive fuzzy control (AFC) algorithms to cope with the challenges of input saturation, external disturbances, and strong uncertainty. For example, a novel adaptive fuzzy control algorithm has been successfully developed that significantly reduces the model complexity by simplifying the parameter settings of the traditional fuzzy methods, and the operational efficiency of the control system is improved [11]. Another important achievement is the proposal of the model-referenced adaptive robust fuzzy control algorithm, which is designed to solve the uncertain nonlinear system in ship helm autopilot, and utilizes the fuzzy logic system to accurately approximate the unknown nonlinear function, which ensures the stable and reliable operation of the ship in the complex environment [38]. In terms of constraints of ship heading, scholars have also made significant progress. To address the challenges associated with heading constraints, grounding risk, and maneuvering in restricted waters, such as narrow waterways and port channels, Liu et al. [39] proposed an event-triggered adaptive fuzzy control strategy. This approach leverages a fuzzy logic system to approximate unknown nonlinear dynamics in the autopilot model, while explicitly incorporating predefined heading constraints into the control architecture. By dynamically adjusting control actions based on event-triggered updates, the method enhances both maneuverability and situational responsiveness, thereby reducing the risk of grounding and improving navigation safety in spatially constrained environments. This study offers a promising solution for safe and reliable autonomous ship operation in confined and complex maritime scenarios. In addition, Tomera [33] explored the application of a fuzzy self-adjusting controller in ship heading control, which combines the advantages of fuzzy logic and linear PID control, enabling adaptive adjustment under varying navigational conditions. While this approach enhances response speed and adaptability, its performance may be less accurate when the rule base or membership functions are poorly designed. This research demonstrates a promising pathway for integrating adaptive fuzzy logic into intelligent maritime control frameworks. To further improve the comprehensive performance of the ship autopilot system, Le et al. [40] proposed a program combining fuzzy control and linear control. This new system not only inherits the stability advantage of PID control, but also incorporates the flexibility and adaptability of fuzzy control, so that the ship can cope with the complex and changeable navigational environment more comfortably. With the continuous deepening of the research on ship motion types and control algorithms, the application value of fuzzy control algorithms in ship motion control has become increasingly prominent. Numerous scholars, such as Wang et al. [41] and Melnyk et al. [42], have comprehensively evaluated the efficiency of fuzzy control algorithms in automatic control systems and thoroughly discussed their practical application effects in ship motion control. These research conclusions not only enrich the application system of fuzzy control theory in the field of ships, but also provide solid theoretical support and a practical basis for the further development of ship autopilot technology in the future. With the continuous deepening of the research on ship motion types and control algorithms, the application value of fuzzy control algorithms in ship motion control is becoming more and more prominent. Numerous researches have verified the remarkable effect of fuzzy control algorithms in enhancing safety, stability, and accuracy of constraints, but also provided solid theoretical support and a practical basis for the further promotion and application of this algorithm in the field of ships.

3.2.3. Neural Network Control Algorithm

As shown in Figure 12, Thanh [43] presented a self-adjusting ship autopilot system based on the concept of neural networks. The system utilized a multilayer feed-forward neural network and was able to adjust the control strategy in real time according to the environmental changes, so as to enable the ship to maintain a stable sailing state in the complex and changing marine environment. Similarly, Pham et al. [25] introduced an intelligent control method designed for ship autopilot systems operating in disturbed environments, which can effectively deal with the effect of external disturbances on the ship’s heading, significantly improve the robustness of the system, and ensure that the ship can still maintain stable navigation under various unfavorable conditions. ANKI et al. [44] applied neural networks as the main controller for berth control for the first time, which laid a solid foundation for subsequent neural network control research and fully demonstrated the great potential of neural networks in complex control tasks. Through the effective processing of the input signal, the neural network can provide precise control instructions during the berthing process, which greatly improves the accuracy of berth control. In the diversified exploration of control technologies, scholars have also proposed control strategies based on fuzzy logic and neural networks. For example, some studies have designed a new ship autopilot control system by combining a bang-bang controller with a fuzzy neural controller [45]. This system fully considers the nonlinear characteristics, which makes the control more accurate and can effectively respond to various complex situations encountered by the ship during navigation. Meanwhile, the application of emerging technologies such as quantum neural network modeling has also attracted extensive attention from scholars. Zhang et al. [46] demonstrated how to construct a quantum neural network model to design a ship steering controller, which significantly improved the convergence performance. The introduction of quantum computing enables the control algorithm to achieve significant improvements in processing speed and computational power, and to better adapt to complex navigation environments. In the research field of autonomous surface ships (MASSs), Ruan et al. [47] delved into the problem of heading control under actuator anomalies, which not only emphasized the technical challenges that may be encountered in practical operation, but also proposed corresponding solutions, further promoting the development of autonomous ship technology. In addition, scholars developed a PID neural network control system (PIDNN) based on an adaptive regulation method. This control system combines the advantages of traditional PID controllers and the adaptive ability of neural networks, which can dynamically adjust the control parameters according to the environmental changes, and significantly improve the response speed and stability of the system.

3.2.4. Adaptive Control Algorithms

Adaptive control algorithms provide an effective solution to the problem of high fuel costs faced by cargo ships and fishing vessels traveling long distances. The algorithm can optimize the control strategy based on real-time data, which not only helps to save fuel and labor costs, but also occupies a pivotal position in the operation [9]. Especially in the variable marine environment, the ship autopilot system can maintain efficient navigational performance and significantly reduce the need for human intervention by applying adaptive control, thus significantly improving the overall automation level. A comprehensive overview of adaptive control under dynamic positioning and marine environmental uncertainties is provided by Sorensen [48], who analyzed the key challenges in convergence, robustness, and implementation for marine applications. To deal with the problem of insufficient heading control accuracy of traditional ship autopilot systems under complex environmental conditions, Amerongen and Job [49,50] proposed a model reference adaptive control (MRAC) method to address the yaw and instability caused by the susceptibility of heading control to external perturbations. MRAC adjusts the actual output of the system in real time by comparing it to the output of the preset reference model and adjusting the controller’s parameters, thus effectively improving the heading control precision of the ship under autopilot state and maintaining the stable navigation of the ship. Aiming to address the problem of insufficient control of complex nonlinear dynamic behaviors that a ship may encounter during navigation, especially the limitations of the control of motion changes caused by environmental factors such as waves and tides, Du et al. and Li et al. [51,52] proposed a robust adaptive nonlinear control algorithm for ship maneuvering. The algorithm aims to effectively cope with the challenges of system uncertainty and unknown input perturbations, and ensure the stability and controllability of the ship under uncertain environments by dynamically adjusting the control signals, which provides one possibility for solving the control difficulties of the ship under complex navigational conditions.
To improve the performance of adaptive control algorithms in ship maneuvering, scholars have introduced several innovative methods. Adaptive dynamic programming (ADP) was adopted as an advanced strategy in the study [53], which can continuously optimize the control strategy by its online learning capability, and flexibly adjust the control parameters in the dynamically changing marine environment. This not only significantly improves the accuracy of heading tracking, but also enhances the robustness of the system. However, traditional PID controllers are often incompetent when facing the velocity-dependent problems. To overcome this limitation, an innovative adaptive PID controller was proposed [9]. This controller is capable of dynamically adjusting its control parameters based on real-time feedback data, which significantly improves the control performance and enables the ship to maintain stable and accurate maneuvering under various speed conditions. In addition, considering that the direct intervention of the crew is indispensable in certain complex or emergency situations, Eraslan et al. [54] have explored the shared control mechanism in depth. This mechanism realizes seamless human–machine collaboration by facilitating information sharing and decision support between the crew and the automated system, which further enhances the ship’s maneuverability and navigational safety. During the continuous optimization of the control algorithm, Chen et al. [55] proposed an improved control algorithm. The algorithm has a solid theoretical foundation of stability and robustness, and its design process is relatively simplified, which can effectively reduce the design complexity. This improvement makes the algorithm more efficient and reliable in practical applications, and injects new vitality into the development of the field of ship adaptive control.

3.2.5. Sliding Mode Control Algorithm

As an efficient nonlinear control strategy, sliding mode control demonstrates significant advantages in dealing with system uncertainties and external disturbances. Its core feature is the design of a specific sliding mode surface and control law, so that the system state can slide smoothly on the sliding mode surface to achieve the expected dynamic performance. Especially in the field of ship automatic control, sliding mode control is favored for its high adaptability to complex dynamic changes and uncertainties. In view of the complex dynamic changes and uncertainties that ships often face in actual operation, scholars have been exploring new sliding mode control methods. Recently, a design of adaptive sliding mode heading keeping and control was proposed by Liu [56], which fully considers the impact of forward speed changes on ship dynamics and can adjust the control parameters in real time to adapt to different sailing conditions. This design significantly enhances the robustness and adaptability of the control system, enabling the ship to maintain a stable heading in a variable environment. In addition, in order to improve the ship’s path tracking performance, scholars proposed a sliding mode control algorithm based on a radial basis function (RBF) neural network [57]. The algorithm utilizes the powerful learning capability of RBF neural networks to optimize the control strategy, thus effectively reducing the tracking error. This advancement not only improves the ship’s navigation ability in complex waterways, but also provides new ideas for the innovation of the sliding mode control algorithm. The actuator saturation problem is an important challenge in the field of ship control. To solve this problem, Ejaz et al. [58] proposed a sliding mode control scheme that considers the actuator saturation characteristics. This scheme realizes effective heading control under actuator saturation conditions by designing suitable sliding mode surfaces and control laws, which not only improves the stability of the system, but also ensures the safety of the ship under extreme conditions. To improve the performance of ship heading control, Chu et al. [59] also explored a sideslip angle compensation method. This method effectively reduces the tracking error by estimating the ship’s side-slip angle in real time and adjusting the control input accordingly. This innovative idea makes the application of sliding mode control in ship automatic control more flexible and versatile, and making it easier to cope with various complex sailing conditions. In addition to the above advances, an iterative sliding mode control method based on a differential evolutionary algorithm (DEA) has been proposed to solve the path tracking problem of three-degree-of-freedom ships [60]. This method significantly enhances the performance and efficiency of the control system by optimizing the control parameters through a genetic algorithm. This research combines the optimization algorithm with sliding mode control, which brings a new breakthrough in the field of ship automatic control. In the study of ship maneuvering capability, Wang et al. [61] used the Nomoto model to model the maneuvering of self-propelled ships and designed a closed-loop control strategy. This model can accurately describe the dynamic behavior of the ship in various sailing states, which provides a solid theoretical foundation and practical basis for the design of sliding mode control.

3.2.6. Model Predictive Control Algorithm

In recent years, many studies have focused on improving the efficiency of MPC design. Wang et al. [18] proposed an efficient MPC design based on a projection neural network for the ship autopilot system. This method models for the system dynamics through neural networks, which improves the computational efficiency and control accuracy. The method shows good performance in practical applications and can respond quickly to the dynamic changes of the ship. Aiming at the shortcomings of traditional control methods in coping with external disturbances and model uncertainty, Fadhok et al. [62] proposed a linear manipulation model of the integrated nonlinear MPC. This method significantly improves the robustness of the system by optimizing the control performance of the ship under input constraints, and effectively addresses the challenges posed by the complex marine environment and uncertainty of models. Jin et al. [63] verified the ability of MPC to resist external disturbances through sea experiments and designed a heading controller directly based on MPC. These studies show that MPC can effectively resist environmental changes and improve the ship’s navigation stability. In addition, to handle nonlinear characteristics in ship motion control, Zhang et al. [64] proposed an integrated nonlinear MPC based on a linear maneuvering model. This approach not only combines the simplicity of the linear model, but also fully takes into account the nonlinear characteristics of the ship, which realizes the dual goals of time-optimal planning and disturbance-rejection tracking. In terms of learning control methods, MPC reinforcement learning based on a Gaussian process model (GP-MPC-RL) has gradually emerged in recent years. This method can achieve efficient control in uncertain environments by utilizing the dynamic model of the iterative learning system, which shows excellent adaptive ability and learning potential [65]. To further improve the learning efficiency and sample utilization, Cui et al. [65] introduced sample-efficient probabilistic MPC (SPMPC). By using SPMPC, the scholars can learn and optimize the control strategy with less sample data effectively, which accelerated the process of applying learning control methods in ship autopilot. To verify the effectiveness of MPC control strategies, many studies have used simulation studies. By performing simulation tests on ship autopilot models for path tracking tasks, scholars evaluate the performance of different MPC strategies. For example, Helling et al. [66] demonstrated the application of MPC in complex channel environments, and the results showed that MPC was able to improve the accuracy and stability of a ship’s path tracking significantly. For ship path tracking, some scholars developed practical MPC controllers and verified their effectiveness by experience. Xu et al. [67] demonstrated the performance of MPC under different navigational conditions, showing its advantages in dealing with complex shipping lanes and environments.

3.2.7. Control Algorithms Based on Artificial Intelligence

Liu et al. [68] investigated several deep learning and machine learning techniques to enhance the ship’s ability to follow its course and control its heading in challenging circumstances. Ship trajectory prediction using Automatic Identification System (AIS) data has attracted a lot of attention in recent years. Studies have focused on comparing the latest ship trajectory prediction algorithms, including traditional machine learning and emerging deep learning methods. By analyzing a large amount of historical data, it is possible to build more accurate models to predict the future position and heading of a ship, thus providing support for heading control [69]. Reinforcement learning (RL) algorithms, especially those based on deep Q-networks (DQNs), are applied to path tracking and heading control of ships. The core of this approach lies in learning optimal strategies through interaction with the environment, enabling ships to navigate effectively in calm water and wave conditions. Sivaraj et al. [70,71] demonstrated that the Deep Q-Network (DQN) algorithm enables real-time adaptive heading adjustment under dynamically changing maritime conditions. This aligns with its successful implementation in autonomous driving scenarios, as illustrated in the research by Kendall et al. [72]. Furthermore, for continuous control problems commonly encountered in ship dynamics, deep reinforcement learning techniques, such as the Deep Deterministic Policy Gradient (DDPG) algorithm proposed by Lillicrap et al. [73], offer significant potential to enhance control precision and adaptability. Another important technique is the application of artificial neural networks (ANNs). In ship heading control systems, scholars have explored how multilayer feed-forward neural networks can enhance the effectiveness of heading control. By learning from the input data, the neural network can identify complex nonlinear relationships and thus optimize the control strategy. Le [74] showed that control systems with neural networks exhibit higher robustness and adaptability in dynamic environments. Comparative studies between traditional control algorithms and deep learning algorithms have also gradually increased in ship autonomy research. Wu et al. [75] stated that although traditional algorithms are still effective in some situations, deep learning algorithms show better performance when dealing with large-scale data and complex environments. This comparison provides important guidance for future research, emphasizing the need to combine traditional and modern techniques for more efficient heading control. In addition, machine learning algorithms play an important role in improving the energy efficiency of ships. By optimizing routes and control modes, energy consumption and emissions are significantly reduced, which not only contributes to economic efficiency, but also has a positive significance for environmental protection. Relevant studies have shown that ships with machine learning techniques exhibit lower fuel consumption rates in actual voyages [76].
As shown in Table 1, the advantages and disadvantages of different algorithms for automatic ship rudders are analyzed and discussed.

4. Core Functions

4.1. Heading Control

In the process of ship navigation, the precise control of heading is crucial to ensure navigation safety and efficiency. As shown in Figure 13, accurate heading control is not only the key to navigational efficiency, it is also the cornerstone of crew safety and cargo integrity [77]. Especially in the face of complex sea conditions or crowded harbors, the autopilot, with its excellent adaptability, effectively reduces the crew’s workload and improves the safety and efficiency of navigation significantly [77]. However, the practical application of heading control technology still faces many challenges, such as heading drift due to wave influence [78], real-time changes in environmental factors, and the high sensitivity of system stability to over-adjustment. In order to overcome these challenges, modern techniques of ship heading control have evolved and introduced intelligent control algorithms, such as fuzzy control and neural networks, which are able to handle complex nonlinear dynamic systems more flexibly [79]. At the same time, the trend of integrated systems will promote the in-depth integration of heading control with other navigation and ship control systems to realize the comprehensive intelligence of ship navigation [80]. The optimization of heading control systems focuses on designing accurate control modes. Min et al. [81] proposed an integrated approach to enhance the heading control of ships by improving the timeliness and robustness of navigation. Kang et al. [82] explored a heading control approach utilizing global path-tracking constraints. This approach aims to balance the relationship between navigation efficiency and energy consumption, and by optimizing the path tracking algorithm, it can improve the accuracy of heading control while reducing energy consumption. The complexity of the flow field is an important challenge in the study of heading control. To deal with this issue, Lim et al. [53] explored the method of establishing an accurate mathematical model in the hovering state. As the dynamic change of the flow field makes the heading control more complicated, the responsiveness to the change of the flow field is enhanced by improving the mathematical model, which improves the accuracy of the heading control. The improvement of heading control not only affects the performance underway, but also has a direct positive impact on operations such as mooring and towing. Zhu et al. [1] showed that by improving heading control, the minimum towing moment value of the mooring line can be reduced by 8.2%, thus improving the ship’s ability to stay at a fixed point, which suggests that optimized heading control can significantly improve the ship’s mooring performance.

4.2. Wave Compensation

Modern ship introduces an advanced fluctuation compensation system in response to the shortcomings of traditional ships that are vulnerable to wave impacts in complex marine environments. This system highly relies on sophisticated sensor data and complex control algorithms to achieve real-time monitoring and prediction of wave impacts. By utilizing advanced techniques such as response amplitude operator (RAO) and linear wave approximation, the system can predict the dynamic impacts of waves on the ship accurately and make instantaneous adjustments accordingly, which significantly improves the adaptability of the ship [83]. However, under low-frequency and heavy-load conditions, the dynamic characteristics of the Stewart platform, a key piece of equipment on the ship, poses an important challenge to the control performance of the wave compensation system. To overcome this challenge, Chen et al. [84] specifically considered the dynamic characteristics of the actuator in the design to ensure that the system can quickly respond to and effectively compensate for wave-induced fluctuations, thus guaranteeing the stable operation of the ship. Further, to address the problem that the first-order system in wave compensation is often limited by the lag in the dynamic response of the thruster, Cai et al. [85] and Xu et al. [86] proposed an innovative and adaptive robust dual-loop control scheme. The scheme cleverly integrates the velocity feed-forward loop and the motion control loop; this counteracts the adverse effects of waves on the ship effectively by optimizing the control strategy, which significantly improves the safety and stability of the ship’s operation under harsh sea conditions. Wen et al. [87] focused on the direct solution to the thrust response lag, and they adopted the first-order wave feed-forward compensation method based on the prediction of wave-induced force, which improves the response speed of the system, further makes up for the shortcomings of the traditional first-order fluctuation compensation system, and provides strong technical support for the ship to realize efficient and stable navigation in the complex and changeable marine environment.

4.3. Ship Collision Avoidance

As shown in Figure 14, the first step in the ship navigation and collision avoidance system is information acquisition, which is mainly realized through the automatic radar target analysis (ARPA) system. This system can monitor the surrounding navigation environment in real time and accurately capture and process key data such as the position, speed and heading of other ships. Xue et al. [88] utilized advanced genetic algorithm (GA) technology to evaluate current complex navigation conditions in depth, especially when facing the complex scenario of multi-vessel encounter; the GA can process a variety of variables and select the optimal collision avoidance strategy for the system. In the complex marine environment of multi-ship encounters, the primary task of a ship is to assess the current relative positions and speeds of all target ships accurately [89,90]. In order to achieve this goal, the system incorporates nonlinear heading adjustment technology to adapt to changing navigation conditions by adjusting the speed vector of its own vessel (OS) flexibly [91]. Meanwhile, the system further utilizes an advanced trajectory prediction function, which not only tracks the current navigational state closely, but also predicts the possible future navigational paths prospectively, including the movement trends of obstacles and target vessels, to ensure navigational safety by adhering to the international Collision Avoidance Rules at Sea (COLREGS). In order to achieve efficient collision avoidance, Huang et al. [92] proposed the dynamic control behavior with the system parameterizes, and the process core is the fine-tuning of two key parameters: a well-timed offset of the steering heading angle, and a flexible variation of the propulsion command which covers a wide range of adjustments, from normal speed to full reversal. This highly flexible and responsive control strategy enables the system to adapt to changes in the maritime environment, ensuring that it can accurately judge the safe passage possibilities of all target vessels and plan the optimal routes accordingly. Particularly for autonomous vessels with dynamic restrictions, the kinematic-based K-Reachability Region (KKR) approach shows promise in forecasting practical avoidance maneuvers in multi-vessel scenarios [93]. In multi-vessel scenarios, new methods, such as the Kinematic-based K-Reachability Region (KKR), indicate promise in anticipating workable avoidance tactics. The KKR technique uses forward reachability analysis, as opposed to traditional rule-based or CPA/TCPA-based methods, to identify a collection of all acceptable positions a vessel can reach within a specified time horizon while taking into account its kinematic restrictions. In situations when speed, heading, and turning radius restrictions are crucial, our approach enables a more thorough and constraint-aware evaluation of the accessible region. Bai et al. [94] has demonstrated that KKR-based collision avoidance frameworks can be effectively integrated with COLREGs compliance rules, and even combined with trajectory optimization or reinforcement learning to enhance decision-making flexibility and situational awareness in congested waters [95]. In multi-agent simulations, KKR approaches have shown superior performance in predicting and coordinating safe maneuvers across multiple autonomous surface vessels under real-time constraints [96]. These advancements indicate that KKR is not only theoretically robust but also practically viable for next-generation MASS collision avoidance systems. In addition, optimization-based approaches such as Model Predictive Control (MPC) have been successfully applied to real-time path planning for autonomous vehicles [97]. Furthermore, bio-inspired algorithms like Ant Colony Optimization (ACO) have shown effectiveness in maritime path planning by balancing route efficiency and obstacle avoidance in uncertain environments [98]. In summary, this series of well-designed steps and strategies constitute an intelligent collision avoidance and route optimization solution for ships in multi-ship encounters.

4.4. Safety Control

The safety control technology of ship autopilot plays a vital role in the shipping industry, and with the continuous improvement of ship automation level, the importance of safety control technology becomes more and more significant. In order to monitor the indicators of the autopilot in real time and detect potential problems in time, Chaal et al. [99] proposed a machine learning-based fault prediction model to identify potential faults by analyzing historical data. To improve the adaptability of the autopilot in complex sea conditions, Roberts [100] improved sensors which can monitor the marine environment in real time and adjust the control strategy to enhance the system’s responsiveness to environmental changes. The full utilization of sensor information has become an indispensable part of the realization of safe collision avoidance with automatic rudder on ships. In order to ensure compliance with the International Collision Avoidance Rules at Sea (COLREGS) while minimizing the risk, the collision avoidance control algorithm must deeply integrate and accurately analyze real-time data from various sensors, which can support the formulation of automatic rudder control decisions [101]. Further, in the area of risk assessment and path optimization of automatic ship rudders, Lee et al. [102] proposed RA4MAIS (Risk Assessment Methodology for AI-based Safety). The method is specifically designed to identify risks associated with AI, and by combining machine learning and data analytics techniques, RA4MAIS can identify potential safety hazards in advance so that appropriate preventive measures can be taken. In addition, intelligent algorithms have been increasingly used in collision avoidance systems, in which a variety of techniques including evolutionary algorithms, fuzzy logic, expert systems, and neural networks have been skillfully fused to form powerful hybrid systems [13]. The addition of these intelligent algorithms greatly enhances the decision-making intelligence and adaptability of the ship’s automatic rudder system in complex environments, providing more solid technical support for autonomous collision avoidance of ships. To solve the global path generation and dynamic collision avoidance problems, Thomas et al. [103] proposed a safety-enhanced path planning technique. The technique is divided into two stages: global guidance and local guidance to ensure the safety and effectiveness of the ship during navigation. For the existing safety technology of ship automatic steering, scholars are developing corresponding classification standards. These standards not only focus on the reliability of the technology, but also consider the environmental adaptability and operational safety, which provides a specification for the popularization and application of ship automatic steering technology [104].

4.5. Path Planning

Real-time path planning technology for ship autopilot systems was developed and proposed by Wang et al. [105] and others in order to address the inadequacies of the contemporary shipping industry with regard to the safety and effectiveness of ship navigation in a complex and dynamic marine environment. By dynamically modifying the sailing routes in accordance with the real-time data, this technique effectively addresses the issue of static (such as islands, reefs) and dynamic (such as other ships, ice floes) obstacles that a ship may encounter during navigation. This greatly increases navigation safety, conserves energy, and lessens the crew’s workload. In order to address the issue of excessive operating costs in the modern shipping sector, Vagale et al. [106] stated that the ongoing development of real-time course planning technology offers a fresh approach to this problem as autonomous ships and intelligent ship technology become more and more common. Because real-time path planning technology can dynamically alter sailing routes based on real-time data, avoid superfluous sailing obstacles, and optimize sailing paths, they emphasized that effective path planning may significantly reduce sailing time and fuel consumption.
Zhu et al. [107] proposed a series of strategies to optimize the algorithms aiming at the shortcomings of real-time path planning algorithms in terms of computational efficiency. These strategies not only focus on improving the computational speed of the algorithm to ensure that the sailing routes can be updated quickly and accurately based on real-time data, but also focus on enhancing the reliability of the algorithm, so that it can still work stably, thus meeting the high standard of the modern shipping industry’s need for rapid computation and updating of sailing routes. Scholars such as Vagale et al. [106] have developed an in-depth outlook on the intelligence and adaptive capability of future route planning systems. They stated that the future path planning system should be able to make full use of historical data and real-time environmental information, and make adaptive adjustments through intelligent algorithms, so as to further improve the flexibility and accuracy of path planning. In the complex scenario of a multi-ship encounter, collision avoidance has always been a major challenge in the field of shipping safety. In order to effectively deal with this problem, Zhu et al. [107] proposed an autonomous path planning model that was innovatively based on deep reinforcement learning. With its great learning ability and adaptability, this model can plan optimal navigation paths in changing environments, thus reducing the risk of multi-ship collisions and providing a solid guarantee for shipping safety. Finally, for the overall needs of autonomous ship navigation, Guo et al. [108] proposed the concept of an integrated system based on in-depth research. This system forms a comprehensive and efficient solution for autonomous ship navigation by fusing and integrating path planning with other navigation control systems. This solution not only improves navigation safety and efficiency, but also reduces the crew’s work intensity, providing strong technical support for the intelligent transformation of the modern shipping industry. With the continuous progress of technology and application expansion, this integrated system is expected to become the mainstream development direction in the field of autonomous ship navigation in the future.

4.6. Tracking Control

In recent years, the research on path planning and tracking control of ship autopilot systems has made remarkable progress. In order to improve the navigation accuracy and operational stability of ships, scholars have explored a variety of advanced control methods, including fuzzy control, sliding mode control and neural network control
Fuzzy control is a control method based on fuzzy logic that can handle uncertainty and ambiguity. In ship autopilot, fuzzy control realizes the control of complex systems by transforming the experience of experts into rules. The advantage of fuzzy control is its low dependence on models and its ability to adapt to the variable environment of ships in a complex environment. Vweza et al. [109] proposed a terminal-based sliding mode fuzzy control method, which can eliminate the tracking error in nonlinear ship autopilot systems. By designing appropriate fuzzy control rules, the system can maintain stability in the face of external disturbances. Although fuzzy control performs well in many applications, its design process still relies on expert knowledge and may face the problem of response speed when real-time requirements are high.
Sliding mode control is a robust control technique that is widely used in the control of nonlinear systems. The basic idea is to design a sliding mode surface so that the system state slides on the surface, thus realizing a strong control of the system dynamics. Sliding mode control is especially suitable for dealing with uncertainties and external disturbances in ship motion. Advantages: The sliding mode control has strong disturbance-rejection ability and robustness, and can maintain a good control effect under the change of system parameters or the effect of external disturbances [110]. For example, Li et al. [111] used a sliding mode control algorithm and a finite time extended state observer (FTESO) to solve the curved path problem of an unmanned surface vehicle, and this approach can effectively deal with the model uncertainty and sideslip angle problems. Future research may focus on combining sliding mode control with other intelligent control methods, such as adaptive control and fuzzy control, to further enhance the adaptive capability and control accuracy of the system.
Neural network control is a method of modeling and controlling a system using artificial neural networks. Through learning and training, neural networks can capture the dynamic characteristics of complex systems, thus enabling more precise and adaptive control. Traditional PID controllers provide reliable performance but require extensive manual tuning and may struggle with highly nonlinear ship dynamics. Similarly, fuzzy controllers can effectively handle uncertainties but rely on predefined rule sets that may not generalize well across different operating conditions. In contrast, neural network-based controllers can autonomously learn optimal control strategies by analyzing real-time data, allowing for better compensation of environmental disturbances such as wind and waves. However, the effectiveness of neural networks depends on the quality of training data and requires significant computational resources, making them more suitable for modern intelligent ship systems equipped with high-performance computing capabilities. In recent years, with the development of deep learning technology, the application of neural network control in ship autopilot has gradually increased. Hasanvand et al. [112] proposed an optimal tracking control method based on reinforcement learning which can be effectively applied to unknown unmanned surface vehicles. This method can adapt to the dynamic environment and improve the accuracy of path tracking by continuously learning and optimizing the decision-making process.
Predictive control is a model-based control method that optimizes control inputs by predicting future system behavior. In ship autopilot, model predictive control (MPC) is widely used for heading control and path planning. MPC optimizes the control inputs in real time so that the ship’s trajectory is as close as possible to the intended path. Pan [113] proposed an efficient MPC design methods based on projection neural networks which can effectively cope with the large inertia, strong nonlinearity and large delay characteristics of ships at sea. With the improvement of computational power, MPC will be more widely used in complex environments, especially in dynamic scenarios that require fast response and high-precision control.

4.7. Evaluation Methods

Despite the significant progress in the research and development and application of ship automatic rudder technology, research in the field of its system performance assessment is still lagging. Currently, there is a lack of comprehensive assessment research on the performance of ship automatic rudder systems, which directly leads to the lack of a comprehensive and authoritative assessment system that is widely accepted. Most of the innovative R&D and testing of automatic rudder technology has relied on simulation software platforms, and although they show a high degree of convenience and efficiency at the initial stage, they may overlook the complex and changing factors in the actual marine environment and their potential impact on the performance of automatic rudders. In order to cope with these difficulties, Table 2 discusses several methods of automatic rudder performance evaluation, and it is now necessary to establish a set of scientific and perfect performance evaluation systems and standards for the ship automatic rudder systems. This will ensure the reliability and practicability of the technology and allow for sufficient field verification that the development of a more mature and efficient ship automatic rudder technology can be promoted.

5. Existing Challenges

Ship autopilot systems face multiple challenges in practical applications, and foremost among them is the parameter uncertainty problem. This problem stems from the variable marine environment, unpredictable climate change, and the complex dynamics of the ship itself. These uncertainties act together to cause the system to suffer from both structural and non-structural difficulties in the control process, thus affecting the performance and reliability of the system [119]. In addition, although artificial intelligence (AI) and deep learning technologies have revolutionized the prospects of ship autopiloting, current applications are still in their infancy. Many systems struggle to effectively respond to complex navigational environments and unexpected conditions due to the lack of advanced sensors and efficient algorithmic support. There are also significant challenges in human–machine collaboration. Although autopilot systems are designed to reduce the burden on the crew, they still require crew intervention in emergency situations. This complex collaboration model may increase operational risk due to delayed response [120], and may also erode the crew’s trust in the system if the system design fails to adequately account for human–machine interactions, which in turn may affect overall safety. Finally, the lack of technical standards and regulations has become another major bottleneck restricting the development of the technology. The rapidly developing autonomous driving technology is not yet fully covered by existing shipping regulations. The lack of uniform standards and safety norms not only hinders the popularization of the new technology, but may also hide safety concerns, so there is an urgent need for the relevant institutions to speed up the formulation and improvement of relevant laws and regulations to ensure the healthy development of ship autopilot technology.
As a key technology of modern marine transportation, the core of the ship autopilot system lies in the precise control of heading to enhance the safety and efficiency of navigation. However, environmental factors pose significant challenges to its performance. Wind, waves, tides and other natural conditions such as currents, temperature and visibility have a profound impact on autopilot systems. Strong winds not only cause ships to deviate from their routes, but also reduce heading stability, increase rudder response time, and test the system’s ability to adjust dynamically. Wave fluctuations cause the ship to sway, requiring the autopilot system to perform more frequent and precise control to maintain course stability and avoid safety hazards. Tidal changes affect the ship’s draught depth and stability, especially in harbors and shallow water areas, and the autopilot system needs to flexibly respond to different water conditions. In addition, current speed, temperature fluctuations, and reduced visibility may affect the accuracy of sensors and the decision-making efficiency of the autopilot system. In order to cope with these complex and changing environmental factors, modern ship autopilot systems are gradually moving towards intelligence and adaptive technologies. With the help of cutting-edge technologies such as machine learning and neural networks, the system can analyze environmental changes in real time and adjust the control strategy accordingly. In particular, the data-driven MPC method demonstrates significant advantages in optimizing rudder angle adjustment and improving heading by maintaining accuracy in dynamic environments. Therefore, when designing and optimizing the autopilot system, comprehensively considering and effectively responding to the influence of environmental factors is of inestimable value in enhancing the system’s responsiveness and stability, and safeguarding navigation safety.

6. Future Development Trend

(1)
In terms of application prospects, the future development of ship automatic rudder technology will focus on automation, autonomous decision-making, multi-sensor fusion, navigation optimization, safety enhancement and other research directions. With the continuous progress of technology and cost reduction, automatic rudders will be widely used in all kinds of ships, including ocean-going transportation ships, inland waterway transportation ships, special ships and marine engineering equipment. In the future, automation technology will play an important role in the fields of sea rescue, hazardous environment operations, ocean exploration, etc. It will also promote the change of sea logistics and military applications, and improve the overall operational efficiency and safety.
(2)
In terms of technical prospects, the future development of ship automatic rudder technology will focus on the deepening application of intelligent and autonomous technology, the fusion of high-precision positioning and navigation technology, the continuous optimization of control algorithms and path planning, and the continuous progress of multi-sensor fusion technology. The development and application of these technologies will further enhance the navigation efficiency and safety of ships, and promote the transformation and upgrading of the shipping industry and sustainable development.
(3)
Beyond control algorithm advancements, future research should also address the hydrodynamic efficiency and design optimization of the rudder itself, which plays a vital role in steering performance and energy consumption. Various efficient rudder configurations—such as flap rudders, high-lift rudders, and twisted rudders—have demonstrated significant improvements in lift-to-drag ratios, maneuverability, and propulsion interaction. Flap rudders, for instance, incorporate a trailing edge flap to increase lift at low angles, making them particularly effective in port operations and sharp turning maneuvers [121,122]. Twisted rudders are geometrically adapted to align with the propeller wake, thereby minimizing rotational losses and enhancing overall propulsion efficiency [123,124]. These physical improvements reduce the required rudder angle deflection to achieve equivalent turning moments, thus lowering actuation energy and improving responsiveness. For autonomous and intelligent navigation systems, integrating hydrodynamically optimized rudder designs with adaptive control algorithms could provide synergistic benefits, enhancing not only navigation accuracy but also energy efficiency in real-time. Future development should therefore promote a co-design paradigm, where mechanical rudder characteristics and intelligent control logic are developed in tandem.
(4)
In terms of laws and regulations, with the continuous development of automated rudder technology on ships, relevant laws and regulations will face new challenges focusing on navigation safety, liability determination, data protection, and the development of international standards. The future legal framework needs to clarify the safety standards, emergency response procedures and fault-handling mechanisms of automatic rudder systems to ensure the stability and safety of ships in automated navigation. Meanwhile, data privacy and cybersecurity will become key issues, especially in the context of connecting ships’ automated rudders with other intelligent systems. The issue of liability determination will also become complex, involving multiple parties such as shipowners, technology providers and developers of automated systems.
(5)
Looking ahead, future developments in ship automatic rudder systems must account for the challenges posed by maneuvering in restricted waters and preventing grounding incidents. While autonomous navigation systems are progressing rapidly, navigating in narrow or congested waters such as ports, harbors, and shallow coastal regions presents unique difficulties that traditional systems may not address effectively [125]. These areas are characterized by limited maneuvering space, dynamic environmental changes, and the presence of multiple moving and stationary obstacles. Ensuring safe navigation in such environments requires real-time decision-making, precise path planning, and the integration of advanced sensors for obstacle detection and avoidance [97,98]. Additionally, grounding risk, particularly in areas with variable tides or poorly mapped seafloors, remains a critical concern for autonomous vessels. Research in this area could explore the integration of high-resolution bathymetric mapping, dynamic tide prediction systems, and intelligent control algorithms that allow vessels to adjust their course and speed proactively [126]. Enhancing the ability of automatic rudder systems to handle these complex scenarios will be essential for the safe operation of autonomous ships, especially in confined or hazardous waters.

7. Conclusions

Through a thorough analysis of ship automatic rudder technology, this paper establishes a strong theoretical foundation for this field. It covers a wide range of fundamental subjects, including advanced sensing technology, comprehensive analysis of literature data sets, automatic rudder control algorithms, core functions, current challenges, and future development trends. Through a thorough review of the literature, this paper explains the general direction of ship automatic rudder technology development and delves into great detail about the fundamental role that sensor technology plays in improving navigation stability, accuracy, and safety. The paper summarizes the research progress of automatic rudder control algorithms at home and abroad, and analyzes the applicability and effectiveness of different control strategies under complex sea conditions. This paper focuses on the future development trend of ship autonomous rudder technology and highlights that the key concepts of the future technological innovation will be intelligence, integration, and adaptability. In addition, to encourage the ongoing advancement of related technologies, this paper offers some references and inspiration for the study and development of ship autonomous rudder technology, which contributes to the vigorous development of intelligent ships.

Author Contributions

Conceptualization, M.Z. and Q.C.; methodology, M.Z.; software, D.Y.; validation, M.Z., Q.C., D.Y. and Y.Z.; formal analysis, M.Z.; investigation, Q.C.; resources, Q.C. and Y.Z.; data curation, D.Y.; writing—original draft preparation, Q.C.; writing—review and editing, Q.C.; visualization, Y.Z.; supervision, M.Z.; project administration, M.Z.; funding acquisition, D.Y. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the National Natural Science Foundation of China (Shenhua Yang, No. 52371369), Xiamen Municipal Natural Science Foundation Upper-level Project (Weijun Wang, No. 3502Z202373038), National Key Research and Development Program of China (Weijun Wang, No. 2021YFB390150), Fujian Provincial Natural Science Program (Hongbo Sun, No. 2022J01323, 2023J01325, 2023I0019), Fujian Provincial Science and Technology Program (Weijun Wang, No. 3502ZCQXT2021007), Fujian Provincial Funds for Promoting High-quality Development of Marine and Fishery Industry (Jinhai Chen, No. FJHYF-ZH-2023-10), and Natural Science Foundation of Jimei University (Weijun Wang, No. ZQ2023015).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Zhu, L.; Li, T. Observer-Based Autopilot Heading Finite-Time Control Design for Intelligent Ship with Prescribed Performance. J. Mar. Sci. Eng. 2021, 9, 828. [Google Scholar] [CrossRef]
  2. Zhao, D.; Yang, T.; Ou, W.; Zhou, H. Autopilot Design for Unmanned Surface Vehicle based on CNN and ACO. Int. J. Comput. Commun. Control 2018, 13, 429–439. [Google Scholar] [CrossRef]
  3. Liu, Z.; Lu, X.; Gao, D. Ship Heading Control with Speed Keeping Via a Nonlinear Disturbance Observer. J. Navig. 2019, 72, 1035–1052. [Google Scholar] [CrossRef]
  4. Han, B.; Duan, Z.; Peng, Z.; Chen, Y. A Ship Path Tracking Control Method Using a Fuzzy Control Integrated Line-of-Sight Guidance Law. J. Mar. Sci. Eng. 2024, 12, 586. [Google Scholar] [CrossRef]
  5. Sethuramalingam, T.K. Fuzzy Logic Controller Based Ship Navigation System. J. Phys. Conf. Ser. 2021, 1850, 012113. [Google Scholar] [CrossRef]
  6. Yang, X.; Liu, M.; Wang, X.; Hu, B.; Liu, M.; Wang, X. Ship Network Traffic Engineering Based on Reinforcement Learning. Electronics 2024, 13, 1710. [Google Scholar] [CrossRef]
  7. Pendleton, S.D.; Andersen, H.; Du, X.; Shen, X.; Meghjani, M.; Eng, Y.H.; Rus, D.; Ang, M.H. Perception, Planning, Control, and Coordination for Autonomous Vehicles. Machines 2017, 5, 6. [Google Scholar] [CrossRef]
  8. Wang, C.; Zhang, X.; Li, R.; Dong, P. Path Planning of Maritime Autonomous Surface Ships in Unknown Environment with Reinforcement Learning; Springer: Berlin/Heidelberg, Germany, 2019. [Google Scholar]
  9. Chen, C.; Delefortrie, G.; Lataire, E. Experimental Investigation of Practical Autopilots for Maritime Autonomous Surface Ships in Shallow Water. Ocean Eng. 2020, 218, 108246. [Google Scholar] [CrossRef]
  10. Ridwan, R. Advanced Development of Sensors’ Roles in Maritime-Based Industry and Research: From Field Monitoring to High-Risk Phenomenon Measurement. Appl. Sci. 2021, 11, 3954. [Google Scholar] [CrossRef]
  11. Chen, J.; Shan, Q.; Li, T.; Meng, Y.; Wang, X. Intelligent Ship Autopilot System Control Based on Generalized Fuzzy Hyperbolic Model. In Proceedings of the 2021 International Conference on Security, Pattern Analysis, and Cybernetics (SPAC), Chengdu, China, 18–20 June 2021; pp. 561–565. [Google Scholar]
  12. Innac, A.; Angrisano, A.; Gaglione, S.; Crocetto, N. GPS Precise Positioning Techniques for Remote Marine Applications. In Proceedings of the 2019 IMEKO TC-19 International Workshop on Metrology for the Sea, Genoa, Italy, 3–5 October 2019. [Google Scholar]
  13. Zhang, X.; Wang, C.; Jiang, L.; An, L.; Yang, R. Collision-Avoidance Navigation Systems for Maritime Autonomous Surface Ships: A State of the Art Survey. Ocean Eng. 2021, 235, 109380. [Google Scholar] [CrossRef]
  14. Zhao, W.; Qi, S.; Liu, R.; Zhang, G.; Liu, G. A Review of Underwater Multi-source Positioning and Navigation Technology. In Advances in Guidance, Navigation and Control; Yan, L., Duan, H., Deng, Y., Eds.; ICGNC 2022, Lecture Notes in Electrical Engineering; Springer: Singapore, 2023; Volume 845. [Google Scholar]
  15. Zhang, Z.; Yang, N.W.; Yang, Y.J. Autonomous Navigation and Collision Prediction of Port Channel Based on Computer Vision and Lidar. Sci. Rep. 2024, 14, 11300. [Google Scholar] [CrossRef] [PubMed]
  16. Kim, T.E.; Schröder-Hinrichs, J.U. Research Developments and Debates Regarding Maritime Autonomous Surface Ship (MASS): Status, Challenges and Perspectives; Springer: Berlin/Heidelberg, Germany, 2021. [Google Scholar]
  17. Liang, B.; Zhao, W.; Wang, X.; Wang, X.; Liu, Z. Metric Reliability Analysis of Autonomous Marine LiDAR Systems Under Extreme Wind Loads. J. Mar. Sci. Eng. 2024, 12, 50. [Google Scholar] [CrossRef]
  18. Wang, W.L.C. An Efficient Ship Autopilot Design Using Observer-Based Model Predictive Control. Proc. Inst. Mech. Eng. Part M J. Eng. Marit. Environ. 2021, 235, 203–212. [Google Scholar] [CrossRef]
  19. Zhang, Q.; Wang, L.; Meng, H.; Zhang, W.; Huang, G. A LiDAR Point Clouds Dataset of Ships in a Maritime Environment. IEEE/CAA J. Autom. Sin. 2024, 11, 1681–1694. [Google Scholar] [CrossRef]
  20. Liu, T.; Li, S.; Hu, X.; Guan, H.; Liu, J. Survey on Existing Technologies in Autopilot Systems: Current Status and Future Perspectives. In Proceedings of the 2023 7th International Conference on Transportation Information and Safety (ICTIS), Xi’an, China, 4–6 August 2023; pp. 1409–1416. [Google Scholar]
  21. Beirami, M.; Mousazadeh, H.; Motlagh, A.M. Virtual Rudder Feedback-Based Heading Control in Small Autonomous Surface Vehicles. ISA Trans. 2024, 145, 421–433. [Google Scholar]
  22. Beirami, M.; Lee, H.Y.; Yu, Y.H. Implementation of an Auto-Steering System for Recreational Marine Crafts Using Android Platform and NMEA Network. ISA Trans. 2015, 39, 577–585. [Google Scholar] [CrossRef]
  23. Yue, P.; Liang, X.; Zhao, Y. Adaptive Control for Automatic Rudder Systems Under Varying Sea States and Ship Types. J. Mar. Sci. Technol. 2022, 27, 1552–1563. [Google Scholar]
  24. Zhu, L.; Li, T.; Yu, R.; Wu, Y.; Ning, J. Observer-Based Adaptive Fuzzy Control for Intelligent Ship Autopilot with Input Saturation. Int. J. Fuzzy Syst. 2020, 22, 1416–1429. [Google Scholar] [CrossRef]
  25. Pham, D.-A.; Han, S.-H. Designing a Ship Autopilot System for Operation in a Disturbed Environment Using the Adaptive Neural Fuzzy Inference System. J. Mar. Sci. Eng. 2023, 11, 1262. [Google Scholar] [CrossRef]
  26. Fossen, T.I. Guidance and Control of Ocean Vehicles; John Wiley & Sons: Chichester, UK, 1994. [Google Scholar]
  27. Świder, Z.; Trybus, L.; Stec, A. Consistent Design of PID Controllers for an Autopilot. Pol. Marit. Res. 2023, 30, 78–85. [Google Scholar] [CrossRef]
  28. Moradi, M.H.; Katebi, M.R. Predictive PID Control for Ship Autopilot Design. IFAC Proc. Vol. 2001, 34, 375–380. [Google Scholar] [CrossRef]
  29. Yao, B.; Yang, J.; Zhang, Q.; Guo, Z.; Hu, R. Research and Comparison of Automatic Control Algorithm for Unmanned Ship. In Proceedings of the 3rd International Conference on Control and Robotics Engineering (ICCRE), Nagoya, Japan, 20–23 April 2018; pp. 85–89. [Google Scholar]
  30. Le, M.D.; Nguyen, S.H.; Nguyen, L.A. Study on a New and Effective Fuzzy PID Ship Autopilot. Artif. Life Robot. 2004, 8, 197–201. [Google Scholar] [CrossRef]
  31. Zhang, M.; Chen, M.; Lu, R. An on-Line Adaptive Hybrid PID Autopilot of Ship Heading Control Using Auto-Tuning BP & RBF Neurons. In Proceedings of the 39th Chinese Control Conference (CCC), Shenyang, China, 27–29 July 2020; pp. 3642–3647. [Google Scholar]
  32. Dlaba, T.; Alasan, M.; Krum, M.; Marvui, N. Pso-based Pid Controller Design for Ship Course-keeping Autopilot. Brodogradnja 2019, 70, 4. [Google Scholar] [CrossRef]
  33. Tomera, M. Fuzzy Self-Tuning PID Controller for a Ship Autopilot. In Marine Navigation; CRC Press: Boca Raton, FL, USA, 2017. [Google Scholar]
  34. Kwon, Y.; Park, S.; Lee, J. PID Compensating Model for Design of Ship’s Autopilot. In Proceedings of the 16th International Conference on Control, Automation and Systems (ICCAS), Gyeongju, Republic of Korea, 16–19 October 2016; pp. 1337–1339. [Google Scholar]
  35. Lin, C.F. Robust adaptive fuzzy control for nonlinear systems with unknown dead zones. IEEE Trans. Fuzzy Syst. 2000, 8, 471–482. [Google Scholar]
  36. Zhang, C.; Wan, L.; Liu, Y. Ship Heading Control Based on Fuzzy PID Control; IEEE: New York, NY, USA, 2019. [Google Scholar]
  37. Yang, Y.; Zhou, C.; Ren, J. Model Reference Adaptive Robust Fuzzy Control for Ship Steering Autopilot with Uncertain Nonlinear Systems. Appl. Soft Comput. 2003, 3, 305–316. [Google Scholar] [CrossRef]
  38. Yansheng, Y. Output feedback Adaptive Robust Control Algorithm Applied to Ship Steering Autopilot with Uncertain Non-linear System. In Proceedings of the World Congress on Intelligent Control & Automation, Shanghai, China, 10–14 June 2002; IEEE: New York, NY, USA, 2002. [Google Scholar]
  39. Liu, Y.; Zhu, Q. Adaptive Fuzzy Control for Ship Autopilot with Course Constraint Based on Event-Triggered Mechanism. In Proceedings of the 7th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS), Guangzhou, China, 13–15 November 2020. [Google Scholar]
  40. Le, M.D.; Nguyen, T.H.; Nguyen, T.T.; Nguyen, S.P.; Hoang, T.D. A New and Effective Fuzzy PID Autopilot for Ships. In Proceedings of the 2003 IEEE International Symposium on Computational Intelligence in Robotics and Automation, Kobe, Japan, 16–20 July 2003; IEEE: New York, NY, USA, 2003. [Google Scholar]
  41. Wang, L.; Wu, Q.; Liu, J.; Li, S.; Negenborn, R.R. State-of-the-Art Research on Motion Control of Maritime Autonomous Surface Ships. J. Mar. Sci. Eng. 2019, 7, 438. [Google Scholar] [CrossRef]
  42. Melnyk, O.; Onishchenko, O.; Onyshchenko, S.; Shumylo, O.; Volyanskyy, S.; Bondar, A.; Cheredarchuk, N. Application of Fuzzy Controllers in Automatic Ship Motion Control Systems. Int. J. Electr. Comput. Eng. 2023, 13, 3948–3957. [Google Scholar] [CrossRef]
  43. Thanh, T.L. Self-Tuning Ship Autopilot Based on the Neural Network Concept: An Empirical Approach. Ships Offshore Struct. 2023, 18, 686–694. [Google Scholar]
  44. Im, N.K.; Nguyen, V.S. Artificial Neural Network Controller for Automatic Ship Berthing Using Head-Up Coordinate System. Int. J. Nav. Archit. Ocean Eng. 2018, 10, 235–249. [Google Scholar] [CrossRef]
  45. Kwak, M.K.; Suh, S.H. Design of Neural-Network Based Autopilot Control System(II). J. Soc. Nav. Archit. Korea 1997, 34, 19–26. [Google Scholar]
  46. Zhang, X.; Zhou, H.; Zhao, C.; Su, Z.; Guan, W. Ship Steering Control Based on Quantum Neural Network. Complexity 2019, 2019, 3821048. [Google Scholar]
  47. Ruan, M.; Wang, A.; Wang, D. Shared Control of Ship Autopilots and Human Pilots for Maritime Autonomous Surface Ship in the Presence of Actuator Anomalies. Syst. Sci. Control. Eng. 2022, 10, 300–311. [Google Scholar] [CrossRef]
  48. Sorensen, A.J. A Survey of Dynamic Positioning Control Systems. Annu. Rev. Control 2011, 35, 123–136. [Google Scholar] [CrossRef]
  49. Amerongen, J.V. Adaptive Steering of Ships—A Model Reference Approach. Automatica 1984, 20, 3–14. [Google Scholar] [CrossRef]
  50. Van Amerongen, J. A Model for Reference Adaptive Autopilot Ships. Practical Results. IFAC Proc. Vol. 1981, 14, 1007–1014. [Google Scholar] [CrossRef]
  51. Du, J.; Guo, C. Nonlinear Adaptive Ship Course Tracking Control Based on Backstepping and Nussbaum Gain. In Proceedings of the 2004 American Control Conference, Boston, MA, USA, 30 June 2004–2 July 2004; IEEE: New York, NY, USA, 2005. [Google Scholar]
  52. Li, B.; Liu, S.; Zhang, G.; Zhang, X. Robust Adaptive Course-Keeping Control and Its Application to Ship Autopilot. In Proceedings of the China Automation Congress (CAC), Xiamen, China, 25–27 November 2022; pp. 566–570. [Google Scholar]
  53. Lim, C.C.; Forsythe, W. Autopilot for Ship Control. Part 1: Theoretical Design. IEE Proc. D Control Theory Appl. 1983, 130, 281–287. [Google Scholar] [CrossRef]
  54. Eraslan, E.; Yildiz, Y.; Annaswamy, A.M. Shared Control Between Pilots and Autopilots: Illustration of a Cyber-Physical Human System; IEEE: New York, NY, USA, 2019. [Google Scholar]
  55. Chen, X.; Zhang, X. Nonlinear Feedback Control Based on ANFIS. In Proceedings of the 12th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD), Zhangjiajie, China, 15–17 August 2015; IEEE: New York, NY, USA, 2015. [Google Scholar]
  56. Liu, Z. Adaptive Sliding Mode Control for Ship Autopilot with Speed Keeping. Pol. Marit. Res. 2018, 25, 21–29. [Google Scholar] [CrossRef]
  57. Zhang, H.; Zhang, X.; Bu, R. Radial Basis Function Neural Network Sliding Mode Control for Ship Path Following Based on Position Prediction. J. Mar. Sci. Eng. 2021, 9, 1055. [Google Scholar] [CrossRef]
  58. Ejaz, M.; Chen, M. Sliding Mode Control Design of a Ship Steering Autopilot with Input Saturation. Int. J. Adv. Robot. Syst. 2017, 14, 1729881417703568. [Google Scholar] [CrossRef]
  59. Chu, R.; Liu, Z.; Chu, Z. Improved Super-Twisting Sliding Mode Control for Ship Heading with Sideslip Angle Compensation. Ocean Eng. 2022, 260, 111996. [Google Scholar] [CrossRef]
  60. Yan, H.; Xiao, Y.; Li, Q.; Wang, R. Differential Evolution Algorithm-Based Iterative Sliding Mode Control of Underactuated Ship Motion. Comput. Intell. Neurosci. 2021, 2021, 4675408. [Google Scholar] [CrossRef] [PubMed]
  61. Wang, J.; Wang, J.; Liang, X.; Liu, J.; Cheng, S.; Yi, H. Real-Sea Validation of a Model Predictive Controller’s Inherent Robustness for Medium-Scale Unmanned Trimaran Heading. Ocean Eng. 2024, 313, 119513. [Google Scholar] [CrossRef]
  62. Fadhok, M.I.; Pramujati, B.; Nurhadi, H. Design of sliding mode control for maneuver autonomous surface vehicle using genetic algorithm. AIP Conf. Proc. 2024, 2927, 040009. [Google Scholar]
  63. Jin, Z.; Liu, S.; Jin, L.; Chen, W.; Yang, W. Model Based Robust Predictive Control of Ship Roll/Yaw Motions with Input Constraints. Appl. Sci. 2020, 10, 3377. [Google Scholar] [CrossRef]
  64. Zhang, M.; Hao, S.; Wu, D.; Chen, M.L.; Yuan, Z.M. Time-Optimal Obstacle Avoidance of Autonomous Ship Based on Nonlinear Model Predictive Control. Ocean Eng. 2022, 266, 112591. [Google Scholar] [CrossRef]
  65. Cui, Y.; Osaki, S.; Matsubara, T. Reinforcement Learning Ship Autopilot: Sample-Efficient and Model Predictive Control-Based Approach. In Proceedings of the 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Macau, China, 3–8 November 2019. [Google Scholar]
  66. Helling, S.; Meurer, T. Dual Collision Detection in Model Predictive Control Including Culling Techniques. IEEE Trans. Control Syst. Technol. 2023, 31, 2449–2464. [Google Scholar] [CrossRef]
  67. Xu, P.; Qin, H.; Ma, J.; Deng, Z.; Xue, Y. Data-Driven Model Predictive Control for Ships with Gaussian. Process. Ocean Eng. 2023, 268, 113420. [Google Scholar] [CrossRef]
  68. Liu, J.; Chen, Z.; Zhou, J.; Xue, A.; Peng, D.; Gu, Y.; Chen, H. Research on A Ship Trajectory Classification Method Based on Deep Learning. Chin. J. Inf. Fusion 2024, 1, 3–15. [Google Scholar] [CrossRef]
  69. Li, H.; Jiao, H.; Yang, Z. Ship Trajectory Prediction Based on Machine Learning and Deep Learning: A Systematic Review and Methods Analysis. Eng. Appl. Artif. Intell. 2023, 126, 107062. [Google Scholar] [CrossRef]
  70. Sivaraj, S.; Rajendran, S.; Prasad, L.P. Data Driven Control Based on Deep Q-Network Algorithm for Heading Control and of a in Calm Water and Waves. Ocean Eng. 2022, 259, 111802. [Google Scholar] [CrossRef]
  71. Sivaraj, S.; Rajendran, S. Heading Control of a Ship Based on Deep Reinforcement Learning (RL). In Proceedings of the OCEANS 2022-Chennai, Chennai, India, 21–24 February 2022. [Google Scholar]
  72. Kendall, A.; Hawke, J.; Janz, D.; Mazur, P.; Reda, D.; Allen, J.M. Learning to Drive in a Day. In Proceedings of the 2019 International Conference on Robotics and Automation (ICRA), Montreal, QC, Canada, 20–24 May 2019. [Google Scholar]
  73. Lillicrap, T.P.; Hunt, J.J.; Pritzel, A.; Heess, N.; Erez, T.; Tassa, Y. Continuous control with deep reinforcement learning. arXiv 2015, arXiv:1509.02971. [Google Scholar]
  74. Le, T.T. Ship Heading Control System Using Neural Network. J. Mar. Sci. Technol. 2021, 26, 963–972. [Google Scholar] [CrossRef]
  75. Wu, C.; Yu, W.; Li, G.L.W. Deep Reinforcement Learning with Dynamic Window Approach Based Collision Avoidance Path Planning for Maritime Autonomous Surface Ships. Ocean Eng. 2023, 284, 115208. [Google Scholar] [CrossRef]
  76. Jiang, Z. The Application of Artificial Intelligence Technology in Shipping: A Bibliometric Review. J. Mar. Sci. Eng. 2024, 12, 624. [Google Scholar] [CrossRef]
  77. Asfihani, T.; Chotimah, K.; Fitria, I.; Subchan. Ship Heading Control Using Nonlinear Model Predictive Control. In Proceedings of the 3rd International Seminar on Research of Information Technology and Intelligent Systems (ISRITI), Yogyakarta, Indonesia, 10 December 2020; pp. 306–309. [Google Scholar]
  78. Liu, Z.; Chu, R. Robust Adaptive Heading Control for a Surface Vessel with Drift Angles. Ocean Eng. 2024, 205, 107310. [Google Scholar] [CrossRef]
  79. Shi, X.; Chen, P.; Chen, L. An Integrated Method for Ship Heading Control Using Motion Model Prediction and Fractional Order Proportion Integration Differentiation Controller. J. Mar. Sci. Eng. 2023, 11, 2294. [Google Scholar] [CrossRef]
  80. Yu, P.; Zhou, Y.; Sun, X.; Sang, H.; Zhang, S. Power-Efficient Event-Triggered Adaptive Heading Control for Wave Gliders. Ocean Eng. 2025, 318, 120175. [Google Scholar] [CrossRef]
  81. Min, F.; Yang, F.; Wang, G.; Ye, X. Research on the Heading Control of Underwater Vehicle Under Hover Condition. IEEE Access 2020, 8, 220908–220920. [Google Scholar] [CrossRef]
  82. Kang, H.H.; Lee, D.S.; Lim, J.S.; Lee, S.J.; Jang, J.; Jung, K.H.; Lee, J. A study on the Effectiveness of the Heading Control on the Mooring Line Tension and Position Offset for an Arctic Floating Structure Under Complex Environmental Loads. J. Mar. Sci. Eng. 2021, 9, 102. [Google Scholar] [CrossRef]
  83. Halvorsen, H.S.; Veraas, H.; Landstad, O.; Smines, V.; Fossen, T.I.; Johansen, T.A. Wave Motion Compensation in Dynamic Positioning of Small Autonomous Vessels; Springer: Berlin/Heidelberg, Germany, 2021. [Google Scholar]
  84. Chen, W.; Wang, S.; Li, J.; Lin, C.H.; Yang, Y.; Ren, A. An ADRC-Based Triple-Loop Control Strategy of Ship-Mounted Stewart Platform for Six-DOF Wave Compensation. Mech. Mach. Theory 2023, 184, 105289. [Google Scholar] [CrossRef]
  85. Cai, Y.; Zheng, S.; Liu, W.; Qu, Z.; Han, J. Adaptive Robust Dual-Loop Control Scheme of Ship-Mounted Stewart Platforms for Wave Compensation. Mech. Mach. Theory 2021, 164, 104406. [Google Scholar] [CrossRef]
  86. Xu, Z.; Wang, Z.; Shen, Z.; Sun, Y. Nonlinear Differential and Integral Sliding Mode Control for Wave Compensation System of Ship-Borne Manipulator; SAGE: London, UK, 2021. [Google Scholar]
  87. Wen, Y.; Li, W.; Zhou, S.; Gao, F.; Chen, W. Robust Sliding Mode Control with Adaptive Gravity Estimation of Ship-Borne Stewart Platform for Wave Compensation. Appl. Ocean Res. 2024, 148, 104004. [Google Scholar] [CrossRef]
  88. Xue, Y.; Lee, B.S.; Han, D. Automatic Collision Avoidance of Ships. Proc. Inst. Mech. Eng. M J. Eng. Marit. Environ. 2009, 223, 33–46. [Google Scholar] [CrossRef]
  89. Ito, M.; Zhang, F.; Yoshida, N. Collision Avoidance Control of Ship with Genetic Algorithm. In Proceedings of the IEEE International Conference on Control Applications, Kohala Coast, HI, USA, 22–27 August 1999; IEEE: New York, NY, USA, 1999. [Google Scholar]
  90. Johansen, T.A.; Cristofaro, A.; Perez, T. Ship Collision Avoidance Using Scenario-Based Model Predictive Control. In Proceedings of the IFAC Conference on Control Applications in Marine Systems, Oita, Japan, 13–16 September 2017. [Google Scholar]
  91. Junmin, M.; Mengxia, L.; Weixuan, H.; Xiaohan, Z.; Yixiong, H. Mechanism of Dynamic Automatic Collision Avoidance and the Optimal Route in Multi-Ship Encounter Situations. J. Mar. Sci. Technol. 2020, 26, 141–158. [Google Scholar] [CrossRef]
  92. Huang, Y.; Chen, L.; Negenborn, R.R.; van Gelder, P.H.A.J.M. Collision Avoidance Systems for Maritime Autonomous Surface Ships Considering Uncertainty in Ship Dynamics. IFAC-Pap. 2020, 53, 14614–14619. [Google Scholar] [CrossRef]
  93. Xiao, B.; Zhao, H.; Wang, L. Kinematic-Based K-reachability Region Method for Ship Collision Avoidance Under Dynamic Constraints. Ocean Eng. 2023, 273, 113934. [Google Scholar]
  94. Bai, T.; Xu, Y.; Sun, J. A Colregs-Compliant Collision Avoidance Approach Based on K-Reachability Sets and Trajectory Optimization. IEEE Trans. Intell. Transp. Syst. 2022, 23, 19348–19359. [Google Scholar]
  95. Chen, Y.; Li, X.; Zhang, X. An Improved Reachability-based Method for Autonomous Vessel Collision Avoidance Considering Complex Constraints. J. Navig. 2021, 74, 1285–1304. [Google Scholar]
  96. Nguyen, Q.H.; Yoon, J. Dynamic Kinematic Reachability-Based Trajectory Planning for Multi-agent Autonomous Surface Vehicles. Ocean Eng. 2023, 272, 113855. [Google Scholar]
  97. Kuwata, Y.; Teo, J.; Fiore, G.; Karaman, S.; Frazzoli, E.; How, J.P. Real-Time Motion Planning with Applications to Autonomous Urban Driving. IEEE Trans. Control Syst. Technol. 2009, 17, 1105–1118. [Google Scholar] [CrossRef]
  98. Tam, C.; Bucknall, R. Path Planning of Autonomous Marine Surface Vessels Using Ant Colony Optimisation. J. Navig. 2010, 63, 451–476. [Google Scholar]
  99. Chaal, M.; Ren, X.; BahooToroody, A.; Basnet, S.; Bolbot, V.; Banda, O.V. Research on Risk, Safety, and Reliability of Autonomous Ships: A Bibliometric Review. Saf. Sci. 2023, 167, 106256. [Google Scholar] [CrossRef]
  100. Roberts, G.N. Trends in Marine Control Systems. Annu. Rev. Control 2008, 32, 263–269. [Google Scholar] [CrossRef]
  101. Lyu, H.; Hao, Z.; Li, J.; Li, G.; Sun, X.; Zhang, G.; Yin, Y.; Zhao, Y.; Zhang, L. Ship Autonomous Collision-Avoidance Strategies—A Comprehensive Review. J. Mar. Sci. Eng. 2023, 11, 830. [Google Scholar] [CrossRef]
  102. Lee, C.; Lee, S. A Risk Identification Method for Ensuring AI-Integrated System Safety for Remotely Controlled Ships with Onboard Seafarers. J. Mar. Sci. Eng. 2024, 12, 1778. [Google Scholar] [CrossRef]
  103. Thomas, S.; Gareth, H.; Mcdonald, M.K. Autonomous Ship Collision Avoidance Navigation Concepts, Technologies and Techniques. J. Navig. 2008, 61, 129–142. [Google Scholar]
  104. Guoshuai, L.I.; Zhang, X.; Zhang, A. Research Hotspots and Tendency of Intelligent Ship Berthing Technology. Chin. J. Ship Res. 2024, 19, 3–14. [Google Scholar]
  105. Wang, S.; Huang, M.; Chen, C.; Sun, J.; Ma, F. A Path Planning Method for Ship Collision Avoidance Considering Spatial–Temporal Interaction Effects. Appl. Sci. 2022, 12, 5036. [Google Scholar] [CrossRef]
  106. Vagale, A.; Oucheikh, R.; Bye, R.T.; Osen, O.L.; Fossen, T.I. Path Planning and Collision Avoidance for Autonomous Surface Vehicles I: A Review. J. Mar. Sci. Technol. 2021, 26, 1292–1306. [Google Scholar] [CrossRef]
  107. Zhu, L.; Li, T. Observer-Based Adaptive Fuzzy Prescribed Performance Control for Intelligent Ship Autopilot. Syst. Sci. Control Eng. 2021, 9, 489–496. [Google Scholar] [CrossRef]
  108. Guo, S.; Zhang, X.; Zheng, Y.; Du, Y. An Autonomous Path Planning Model for Unmanned Ships Based on Deep Reinforcement Learning. Sensors 2020, 20, 426. [Google Scholar] [CrossRef] [PubMed]
  109. Vweza, A.O.; Lee, D.J.; Hong, D.P.; Chong, K.T. Fast Terminal Sliding Mode Control for the Ship Steering Problem. Appl. Mech. Mater. 2013, 433–435, 1078–1085. [Google Scholar] [CrossRef]
  110. Londhe, P.S.; Patre, B.M. Adaptive Fuzzy Sliding Mode Control for Robust Trajectory Tracking Control of an Autonomous Underwater Vehicle. Intel. Serv. Robot. 2019, 12, 87–102. [Google Scholar] [CrossRef]
  111. Li, B. New Techniques in Motion Control and Path Planning of Marine Vehicles. J. Mar. Sci. Eng. 2024, 12, 176. [Google Scholar] [CrossRef]
  112. Hasanvand, A.; Seif, M.S. Adaptive Path-Following Control for Displacement Vessels at Any Loading Conditions Under Ocean Disturbances. J. Mar. Sci. Appl. 2024, 23, 900–913. [Google Scholar] [CrossRef]
  113. Pan, Y. Model Predictive Control of Unknown Nonlinear Dynamical Systems Based on Recurrent Neural Networks. IEEE Trans. Ind. Electron. 2012, 59, 3089–3101. [Google Scholar] [CrossRef]
  114. Tomasi, J.; Ro, K. Analytical Modeling and Simulation of an Autonomous Underwater Vehicle with Five Control Surfaces. In Proceedings of the International Society for Optics and Photonics, San Diego, CA, USA, 13–17 August 2006. [Google Scholar]
  115. Han, K.; Sung, J.; Kim, B. Application of an Adaptive Autopilot Design and Stability Analysis to an Anti-Ship Missile. Int. J. Aeronaut. Space Sci. 2011, 12, 78–83. [Google Scholar] [CrossRef]
  116. Raygosa-Barahona, R.; Garcia-Terán, M.Á.; Enriquez, C.; Olguín-Díaz, E. Experimental Evaluation of an Autonomous Surface Craft for Shallow-Water Bathymetry. Mar. Technol. Soc. J. 2017, 51, 59–67. [Google Scholar] [CrossRef]
  117. Belanger, F.; Millan, D.; Cyril, X. Submarine Autopilot Performance Optimization with System Identification. In Proceedings of the ISCSS Conference, Glasgow, UK, 2–4 October 2018. [Google Scholar]
  118. Zhang, P.; Zhu, B.; Zhao, J.; Fan, T.; Sun, Y. Performance Evaluation Method for Automated Driving System in Logical Scenario. Automot. Innov. 2022, 5, 299–310. [Google Scholar] [CrossRef]
  119. Das, S. Applicability of Sliding Mode Control in Autopilot Design for Ship Motion Control. In Proceedings of the International Conference on Recent Advances and Innovations in Engineering (ICRAIE-2014), Jaipur, India, 9–11 May 2014; pp. 1–6. [Google Scholar]
  120. Veitch, E.; Alsos, O.A. A Systematic Review of Human-AI Interaction in Autonomous Ship Systems. Saf. Sci. 2022, 152, 105778. [Google Scholar] [CrossRef]
  121. Kose, K.; Tanaka, H. Analysis of Hydrodynamic Characteristics of High Lift Rudders for Large Ships. Ocean Eng. 2015, 108, 639–649. [Google Scholar]
  122. Andren, H.; Johansson, R.; Svensson, T. On the Performance of High-Lift Rudders for Maneuvering and Energy Efficiency. J. Mar. Sci. Technol. 2019, 24, 756–764. [Google Scholar]
  123. Korkut, E.; Atlar, M. On the Improvement of Propeller–Rudder Interaction by Using a Twisted Rudder. Ocean Eng. 2003, 30, 2275–2293. [Google Scholar]
  124. Molland, A.F.; Turnock, S.R.; Hudson, D.A. Ship Resistance and Propulsion: Practical Estimation of Propulsive Power; Cambridge University Press: Cambridge, UK, 2017. [Google Scholar]
  125. Burmeister, H.C.; Bruhn, W.; Walther, A. Developments Towards a Fully Autonomous Navigation System. TransNav Int. J. Mar. Navig. Saf. Sea Transp. 2014, 8, 61–66. [Google Scholar]
  126. Johansen, T.A.; Perez, T.; Cristofaro, A. Ship Grounding Avoidance Using Predictive Control and Bathymetric Data. Control Eng. Pract. 2016, 53, 146–155. [Google Scholar]
Figure 1. Research architecture diagram.
Figure 1. Research architecture diagram.
Jmse 13 00851 g001
Figure 2. Heat map of keywords.
Figure 2. Heat map of keywords.
Jmse 13 00851 g002
Figure 3. Keywords sorted by prominence intensity.
Figure 3. Keywords sorted by prominence intensity.
Jmse 13 00851 g003
Figure 4. Number of papers published by journal.
Figure 4. Number of papers published by journal.
Jmse 13 00851 g004
Figure 5. Visual chart of keyword.
Figure 5. Visual chart of keyword.
Jmse 13 00851 g005
Figure 6. Visual chart of related technology.
Figure 6. Visual chart of related technology.
Jmse 13 00851 g006
Figure 7. Composition framework diagram of the GNSS.
Figure 7. Composition framework diagram of the GNSS.
Jmse 13 00851 g007
Figure 8. Principle diagram of LIDAR working.
Figure 8. Principle diagram of LIDAR working.
Jmse 13 00851 g008
Figure 9. Rudder angle sensor. (a) Rudder angle sensor outline diagram; (b) Rudder angle sensor internal structure diagram.
Figure 9. Rudder angle sensor. (a) Rudder angle sensor outline diagram; (b) Rudder angle sensor internal structure diagram.
Jmse 13 00851 g009
Figure 10. Schematic diagram of PID control algorithm.
Figure 10. Schematic diagram of PID control algorithm.
Jmse 13 00851 g010
Figure 11. Fuzzy control schematic.
Figure 11. Fuzzy control schematic.
Jmse 13 00851 g011
Figure 12. Graph of adaptive automatic rudder based on BP neural network.
Figure 12. Graph of adaptive automatic rudder based on BP neural network.
Jmse 13 00851 g012
Figure 13. Heading control system structure.
Figure 13. Heading control system structure.
Jmse 13 00851 g013
Figure 14. Flow chart of ship collision avoidance.
Figure 14. Flow chart of ship collision avoidance.
Jmse 13 00851 g014
Table 1. Algorithm comparison of automatic rudder control.
Table 1. Algorithm comparison of automatic rudder control.
Control AlgorithmAdvantages Drawbacks
PID control① Simple structure.
② Strong adaptability and convenient adjustment.
③ High control accuracy, widely used.
④ Model-independent.
① Difficult to adjust parameters.
② Requires retuning in nonlinear or time-varying systems.
③ Limited disturbance-rejection ability.
Fuzzy control① Strong ability to handle uncertainty and nonlinearity.
② Strong robustness to changes in parameters and disturbances of the system.
③ Simple design and high realizability.
① Complicated models and experience-dependent design.
② Performance depends heavily on rule base and membership function design; improper tuning can reduce responsiveness or increase error variance under disturbance.
③ Stability is difficult to prove.
Neural network control① High adaptability to complex and nonlinear ship dynamics, allowing for enhanced accuracy in automatic rudder control, particularly under varying sea conditions.
② Real-time adjustment of the automatic rudder system to achieve a more optimized effect of control.
③ Realize the intelligence and automation of the automatic rudder system, reducing the dependence on manual steering.
① High requirement for training data.
② Long time for training.
③ High computational complexity.
④ Easy overfitting and poor interpretation.
Adaptive control① Effective handling of system uncertainty.
② Strong adaptability to model errors and disturbances.
③ Stronger ability to learn and adapt.
① May be unstable if poorly initialized or with inadequate data.
② Large amount of calculation.
③ System stability depends on the undetermined parameters.
Sliding mode control① Sliding surface control is extremely insensitive to changes in system parameters and external perturbations.
② The sliding surface control can realize fast sliding and accurate tracking of the system state.
③ The algorithm is relatively simple and easy to understand.
① The introduction of high-frequency regulation signals can lead to high-frequency oscillations in the control system.
② The phenomenon of overshooting may occur.
③ Controller output may be jittery in the vicinity of the sliding mold surface.
Model predictive control① Strong disturbance-rejection ability, can tolerate changes in system parameters and external interference to some extent.
② Model predictive control exhibits strong dynamic performance, including rapid response, stability under constraints, and the ability to maintain output tracking under varying conditions.
③ It is easy to deal with problems of multi-constraint optimization.
① Complexity in parameter tuning, including the adjustment of the prediction horizon, control horizon, and weighting matrices.
② Large amount of calculation.
③ High sensitivity to model mismatch.
Artificial Intelligence Control① Adaptive decision-making capability, precision, and accuracy.
② Adaptability and flexibility.
③ Strong scalability and customizability.
① High cost.
② High technical difficulty.
③ Strong dependence and vulnerability.
Table 2. Automatic rudder performance evaluation method.
Table 2. Automatic rudder performance evaluation method.
Years Literatures Evaluation Methods
2006Tomasi J [114]Dynamic modeling and numerical simulation of an autonomous underwater vehicle (AUV) with five hydrodynamic control surfaces are presented.
2011Han K et al. [115]Evaluation of automatic rudders using the stability of neural network control systems is presented.
2017Raygosa-Barahona [116]A simple experimental method for tuning low-level PID control parameters is presented.
2018Belanger et al. [117]Based on techniques such as computer simulation modeling and computational fluid dynamics (CFD), the hydrodynamic coefficients of a ship can be calculated to assess its performance under different sailing conditions.
2020Chen et al. [9]A model of a two-loop control system is proposed to evaluate the effect of external disturbances on a ship’s automatic rudder.
2022Zhang et al. [118]A performance evaluation method for an autonomous driving system in the full parameter space of logical scenarios is proposed.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zheng, M.; Chen, Q.; Yan, D.; Zhang, Y. A Comprehensive Review of Ship Automatic Rudder Technology: Development, Challenges, and Prospects. J. Mar. Sci. Eng. 2025, 13, 851. https://doi.org/10.3390/jmse13050851

AMA Style

Zheng M, Chen Q, Yan D, Zhang Y. A Comprehensive Review of Ship Automatic Rudder Technology: Development, Challenges, and Prospects. Journal of Marine Science and Engineering. 2025; 13(5):851. https://doi.org/10.3390/jmse13050851

Chicago/Turabian Style

Zheng, Minjie, Qianqiang Chen, Daowei Yan, and Yawei Zhang. 2025. "A Comprehensive Review of Ship Automatic Rudder Technology: Development, Challenges, and Prospects" Journal of Marine Science and Engineering 13, no. 5: 851. https://doi.org/10.3390/jmse13050851

APA Style

Zheng, M., Chen, Q., Yan, D., & Zhang, Y. (2025). A Comprehensive Review of Ship Automatic Rudder Technology: Development, Challenges, and Prospects. Journal of Marine Science and Engineering, 13(5), 851. https://doi.org/10.3390/jmse13050851

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop