Special Issue on “Challenges and Advances of Process Control Systems”

1. Introduction

Control systems stand as the crucial nervous system of modern industrial and technological processes, essential for achieving high performance, safety, and efficiency across diverse applications, from complex microgrids to advanced robotics [1,2,3]. However, the effectiveness of these systems is often hampered by inherent real-world complications. Traditional control methods, such as standard Proportional–Integral–Derivative (PID) controllers, frequently fall short when faced with persistent challenges like system parameter uncertainties, non-linear disturbances, and highly variable operating conditions [4,5,6,7]. This performance degradation is critically apparent in high-performance sectors, including electric motor drives and sophisticated bioreactors, where ambiguity can severely compromise stability and reliability [8].

Consequently, the field is undergoing a paradigm shift, rapidly moving towards intelligent and fractional-order control algorithms to overcome these fundamental deficits, thereby prioritizing enhanced robustness, flexibility, and dynamic response [9,10,11,12,13]. This push has yielded powerful hybrid methodologies; for instance, combining classical PI control with cutting-edge machine learning tools like the Deep Q-Network (DQN) creates adaptive, high-accuracy regulators ideal for challenging environments such as DC microgrids [14,15,16,17]. Furthermore, sophisticated tools are vital for modeling physical systems; in applications like steam turbine power generation, detailed dynamic analysis is now indispensable for designing new control strategies that can effectively manage heat extraction and maintain crucial system frequency stability [18,19,20]. The intensive research efforts highlight that the future of control lies in resilient, adaptive, and predictive algorithms that are fully equipped to handle the growing complexity and volatility inherent in 21st-century engineering and technology. The evolution moves beyond merely stabilizing a system; the new goal is intelligent autonomy. This necessitates advanced strategies that can proactively forecast disturbances (such as wind power fluctuations) and dynamically adjust parameters without human intervention or reliance on idealized models. Fractional-order control (FOC) provides enhanced flexibility and fine-tuning capabilities, while the integration of Deep Reinforcement Learning (DRL) allows controllers to learn optimal policies in real-time within uncertain and non-linear environments [21,22]. This synergy promises systems with unprecedented robustness and agility, capable of achieving near-perfect stability and response times while simultaneously adhering to complex operational constraints across diverse fields, from smart storage power systems and industrial extraction heat to precision biotechnology.

2. Current Research

2.1. Dynamic Simulation Model of Single Reheat Steam Turbine and Speed Control System Considering the Impact of Industrial Extraction Heat

This research presents an in-depth analysis of the dynamic characteristics and speed control system of a single reheat steam turbine generator unit operating under variable conditions. A comprehensive simulation model was developed, integrating the speed control system, actuator, turbine body, and once-through boiler dynamic coupling, specifically to evaluate the impact of the heat extraction system on the unit’s behavior.

The core focus was on revealing how the heat extraction regulation process affects key operating parameters and the unit’s crucial system frequency regulation capability. Using the actual parameters of a 300 MW heat unit in Guangxi, the model’s dynamic responses were simulated under typical conditions, including primary frequency regulation and load disturbances.

The simulation results showed excellent agreement with actual engineering data, confirming the model’s ability to accurately characterize the dynamic performance of the heat unit. This validated the dynamic performance of the heat system module and the rationality of its control parameter design. Ultimately, this study offers a reliable model and theoretical basis for accurate dynamic simulation and the development of optimized control strategies for power system frequency regulation.

2.2. The Voltage Regulation of Boost Converters via a Hybrid DQN-PI Control Strategy Under Large-Signal Disturbances

This paper focuses on improving the performance of a DC-DC boost converter, which is vital for connecting low-voltage sources to high-voltage buses in DC microgrids. To enhance dynamic response and robustness against disturbances and parameter variations, this research proposes a hybrid control strategy integrating a classical Proportional–Integral (PI) controller with a Deep Q-Network (DQN) agent.

The hybrid framework strategically combines the PI control’s precision for steady-state regulation with the DQN’s capability to learn optimal control policies in dynamic and uncertain environments. Simulation results, developed in MATLAB 2024/Simulink, validate the efficacy of this approach.

The proposed hybrid controller demonstrated a significantly faster transient response and superior robustness when compared against conventional PI and fuzzy logic controllers. Furthermore, the inclusion of PI-based fine-tuning effectively compensates for the precision limitations often caused by the DQN’s discrete action space, enabling high-accuracy voltage regulation without requiring an explicit system model.

2.3. Multi-Scale Decomposition and Hybrid Deep Learning CEEMDAN-VMD-CNN-BiLSTM Approach for Wind Power Forecasting

This study addresses the challenge of wind power volatility and uncertainty by introducing a powerful hybrid model for forecasting, combining multiple decomposition and deep learning techniques. The goal is to significantly enhance prediction accuracy for improved grid operation.

The methodology begins by utilizing complete ensemble empirical mode decomposition with adaptive noise to break down the original wind power data into multiple intrinsic mode functions (IMFs). To handle complexity, the high-frequency IMFs are then further refined using variational mode decomposition after classification via sample entropy and k-means clustering.

Next, the decomposed signals are fed into a convolutional neural network to effectively extract local spatiotemporal features. This is followed by a bidirectional long short-term memory network, which is used to capture crucial bidirectional temporal dependencies within the complex data.

Experimental results demonstrated the superiority of this hybrid model over several established benchmarks, including ARIMA and various LSTM combinations. The model achieved high accuracy metrics, notably a Root Mean Squared Error of 8.192 and an impressive coefficient of determination (R²) of 0.9840, offering a reliable new solution for precise wind power forecasting.

2.4. Implementation of Fuzzy PID Controller to an Isolated Wind/Battery/Super Magnetic Energy Storage Power System

This work demonstrates an eco-friendly standalone microgrid integrating wind power, a battery-based Energy Storage System (ESS), and a Superconducting Magnetic Energy Storage (SMES) system. To overcome typical microgrid control challenges, a novel Fuzzy Proportional–Integral–Derivative (FPID) controller was introduced. While wind and ESS satisfy normal demand, the SMES system significantly supports transient performance against load fluctuations. Simulation results confirmed that the combination of the FPID controller and SMES provided superior overall microgrid responses, ensuring continuous, regulated AC power despite disturbances. This optimal configuration achieved a substantial 24.4% average reduction in overshoot compared to other tested scenarios, validating its robust stability. Finally, a Hardware-in-the-Loop emulator was utilized to verify these excellent simulation outcomes.

2.5. Optimal Fractional-Order Controller for Fast Torque Response of an Asynchronous Motor

The research addresses the poor performance of traditional PID controllers in Asynchronous Motor (AM) drives utilizing Direct Torque Control, which are common in applications like electric cars. To significantly enhance speed and torque response, the study proposes an optimized AM drive regulated by DTC and employing a Fractional-Order PI (FOPI) controller. The parameters of this novel FOPI controller were precisely tuned using the Particle Swarm Optimization algorithm.

Comparative simulations confirmed that the optimized FOPI controller delivered the best performance against standard PI and optimized PI controllers. The proposed method yielded substantial improvements, resulting in an impressive 84.4% average reduction in settling time and near elimination of the steady-state error. Furthermore, testing under parameter uncertainties demonstrated that the optimized FOPI provides excellent robustness, making it a highly reliable solution for high-performance AM drives.

2.6. Robustness and Scalability of Incomplete Virtual Pheromone Maps for Stigmergic Collective Exploration

This work introduces the Swarm Guiding and Communication System (SGCS), a decision-making and information-sharing framework for robot swarms that operates solely via close-range peer-to-peer communication and requires no centralized control. The system successfully imitates ant colony behavior: each robot makes decisions based on an incomplete virtual pheromone map that is updated only when interacting with another agent.

Unlike similar systems that often rely on unlimited communication range or environmental modification, the SGCS demonstrates remarkable autonomy. A computer simulation was developed to rigorously assess the framework’s effectiveness and robustness in covering an unknown area.

The results proved that the pheromone-based approach outperforms an unbiased random walk in both consistency and the time required to achieve 99% area coverage, regardless of the number of agents deployed. Furthermore, the framework inherently exhibited strong resilience to individual robot failures.

2.7. PID Controller Design for an E. coli Fed-Batch Fermentation Process System Using Chaotic Electromagnetic Field Optimization

The paper focuses on the optimal tuning of a PID controller essential for maintaining a precise glucose concentration set point in an E. coli fed-batch cultivation process. To accurately model the system—including biomass, substrate, and product dynamics—the researchers developed and utilized a novel hybrid metaheuristic technique called Chaotic Electromagnetic Field Optimization (CEFO).

CEFO was created by integrating ten different chaotic maps into the standard Electromagnetic Field Optimization (EFO) algorithm, significantly enhancing its exploration capabilities. This hybridization proved highly effective for both model parameter identification and PID controller tuning.

Compared to the classical EFO algorithm, the proposed CEFO method demonstrated superior performance, achieving a 30% improvement in the objective function. Based on the derived mathematical models, the CEFO-tuned PID controllers exhibited robust control system performance. Specifically, the third PID controller achieved a fast settling time of approximately 9 min with an overshoot of only 15%, validating CEFO as an effective methodology for designing high-quality closed-loop systems in cultivation processes.

2.8. Influence Mechanism of Ambient Air Parameters on the Rotational Stall of an Axial Fan

This study presents a numerical investigation into the performance of a dual-stage axial-flow fan within a power plant setting, utilizing RANS equations coupled with the Realizable k–ε turbulence model. Under normal steady-state conditions, the simulations revealed a positive correlation between the mass flow rate/outlet pressure and gas density, and a negative correlation with dynamic viscosity. Crucially, the volumetric flow rate at the fan’s maximum outlet pressure remained consistent regardless of air density changes.

The research also analyzed the fan’s behavior during stall conditions. The volumetric flow rate was found to decrease rapidly at stall inception before stabilizing to a slower rate of decline. Analysis of the three-dimensional flow field confirmed that stall primarily initiates at the blade tip of the first-stage rotor.

As the flow rate decreases, the tip leakage area expands, causing the leakage vortex trajectory to shift toward the blade’s leading edge. Upon reaching a critical point, the backflow induced by this leakage vortex obstructs the entire blade tip passage, progressively forming a stall cell that simultaneously affects adjacent flow passages.

2.9. An Application of Lean Techniques to Construct an Integrated Management Systems Preventive Action Model and Evaluation: Kaizen Projects

This research focuses on integrating the philosophies of Occupational Health and Safety and Quality Management Systems, both of which enforce a culture of continual improvement toward zero incidents and zero defects. The study addresses the need for robust problem-solving, encompassing quality nonconformances, safety incidents, and engineering breakdowns.

To systematically address these issues, the paper proposes a novel troubleshooting system designed to evaluate continual improvement projects, drawing upon total quality management, lean principles, Kaizen concepts, and quality standards.

The system operates in three phases: fault recording (capturing complete nonconformance details), problem classification and root cause analysis, and finally, database management and project evaluation (applying the Kaizen philosophy to track permanent solutions). When implemented in a case study company, this model significantly upgraded the existing problem-solving structure, leading to drastic improvements and effective evaluation of continual improvement projects.

2.10. Security Assessment of Industrial Control System Applying Reinforcement Learning

This paper investigates the security vulnerabilities of industrial control systems, whose hierarchical structures and communication backbones make them susceptible to cyberattacks. To efficiently expose these systemic flaws, the research employs a reinforcement learning (RL) extended attack graph.

Specifically, the system environment is modeled using the State–Action–Reward–State–Action (SARSA) technique, where the RL agent acts as the attacker. The attacker’s objective is to achieve the highest possible cumulative reward, corresponding to the greatest system damage accomplished with the fewest actions.

The results successfully identified the worst-case assault scheme, which yielded a total reward of 42.9. Crucially, the analysis pinpointed the most severely affected subsystems, providing critical insights for system managers to prioritize mitigation efforts against potential threats.

2.11. CFD Modeling and Experimental Validation of the Flow Processes of an External Gear Pump

This paper details the development and validation of a two-dimensional (2D) Computational Fluid Dynamics (CFD) model designed to analyze the flow processes within a specific external gear pump specimen. The primary goal was to numerically determine the pump’s key characteristics, particularly the flow rate as a function of pressure and time.

A comprehensive numerical study was conducted across 42 distinct operating modes, systematically varying the rotational frequency (950–1450 min⁻¹) and pressure (5–150 bar). To validate these results, a dedicated laboratory experimental setup equipped with a modern data acquisition system was implemented, allowing for testing under identical conditions.

The study also introduced an original methodology within the 2D CFD framework: adjusting the gear face width to account for the influence of the discharge channels on the pump’s displacement volume. Despite the acknowledged limitations of the simplified 2D model, the comparison between numerical and experimental results yielded a very strong match, with the average relative error index ranging remarkably between 93% and 97%.

2.12. Experimental Study of Sound Pressure Level in Hydraulic Power Unit with External Gear Pump

This work details the experimental measurement and analysis of the Sound Pressure Level (SPL) generated by a hydraulic power unit equipped with an external gear pump. The study utilized a specially developed laboratory setup, which mimics common hydraulic power unit architectures, and involved a detailed description of both the hydraulic system and the specialized measuring equipment.

The experimental design included two main configurations and specific operating parameters. The results are presented as magnitude frequency responses and were thoroughly compared using various quantitative indicators. Importantly, parametric models were derived by approximating the experimental data for specific operating modes.

These resultant models are valuable for future research aimed at reducing the SPL through passive methods (like damping rings) or active controls (such as electric motor frequency adjustments). The comprehensive quantitative analysis established in this work provides a reliable baseline for comparison when evaluating the effectiveness of any subsequent noise reduction strategies.

3. Conclusions

The Special Issue “Challenges and Advances of Process Control Systems” brings together the latest scientific research dedicated to exploring both the current theoretical and technological frontiers of control engineering. This collection aims to illuminate critical issues and opportunities arising from the application and development of advanced control algorithms.

The featured algorithms span a wide and sophisticated range of methodologies, including robust, optimal, adaptive, and predictive control, alongside modern techniques such as fuzzy logic and intelligent control.

Beyond theory, the Issue covers practical control system applications across numerous critical industrial sectors. Topics explored include advancements in manufacturing and production processes, power generation and distribution, robotics, hydraulics, and biotechnology.

By examining these diverse applications, we believe that this Special Issue will offer a comprehensive overview of the cutting-edge research being conducted in the area of process control systems. The goal is to highlight key findings and draw attention to emerging trends and activities that promise to define the future of the field.