Search Results (52)

Short-term trading presents a high-dimensional prediction problem, where the profitability of trading signals depends not only on model accuracy but also on how financial labels are defined and aligned with market dynamics. Traditional approaches often apply uniform modeling choices across assets, overlooking the asset-specific nature of volatility, liquidity, and market response. In this work, we introduce a structured, label-aware machine learning pipeline aimed at maximizing short-term trading profitability across four major benchmarks: S&P 500 (SPX), NASDAQ-100 (NDX), Dow Jones Industrial Average (DJI), and the Tadāwul All-Share Index (TASI and twelve of their most actively traded constituents). Our solution systematically evaluates all combinations of six model types (logistic regression, support vector machines, random forest, XGBoost, 1-D CNN, and LSTM), eight look-ahead labeling windows (3 to 10 days), and four feature subset sizes (44, 26, 17, 8 variables) derived through Random Forest permutation-importance ranking. Backtests are conducted using realistic long/flat simulations with zero commission, optimizing for Percentage Profit and Profit Factor on a 2005–2021 train/2022–2024 test split. The central contribution of the framework is a labeling-aware search mechanism that assigns to each asset its optimal combination of model type, look-ahead horizon, and feature subset based on out-of-sample profitability. Empirical results show that while XGBoost performs best on average, CNN and LSTM achieve standout gains on highly volatile tech stocks. The optimal look-ahead window varies by market from 3-day signals on liquid U.S. shares to 6–10-day signals on the less-liquid TASI universe. This joint model–label–feature optimization avoids one-size-fits-all assumptions and yields transferable configurations that cut grid-search cost when deploying from index level to constituent stocks, improving data efficiency, enhancing robustness, and supporting more adaptive portfolio construction in short-horizon trading strategies. Full article

To address the challenges of inefficient path planning, discontinuous trajectories, and inadequate safety margins in autonomous spraying vehicles for greenhouse environments, this paper proposes a hierarchical motion control architecture. At the global path planning level, the heuristic function of the A* algorithm was redesigned to integrate channel width constraints, thereby optimizing node expansion efficiency. A continuous reference path is subsequently generated using a third-order Bézier curve. For local path planning, a state-space sampling method was adopted, incorporating a multi-objective cost function that accounts for collision distance, curvature change rate, and path deviation, enabling the real-time computation of optimal obstacle-avoidance trajectories. At the control level, an adaptive look-ahead distance pure pursuit algorithm was designed for trajectory tracking. The proposed framework was validated through a Simulink-ROS co-simulation environment and deployed on a Huawei MDC300F computing platform for real-world vehicle tests under various operating conditions. Experimental results demonstrated that compared with the baseline methods, the proposed approach improved the planning efficiency by 38.7%, reduced node expansion by 16.93%, shortened the average path length by 6.3%, and decreased the path curvature variation by 65.3%. The algorithm effectively supports dynamic obstacle avoidance, multi-vehicle coordination, and following behaviors in diverse scenarios, offering a robust solution for automation in facility agriculture. Full article

Block-based hybrid video coders typically use inter-prediction and bidirectionally coded (B) frames to improve compression efficiency. For this purpose, they employ look-ahead buffers, perform out-of-sequence frame coding, and implement similarity search-based general-purpose algorithms for motion estimation. While effective, these methods increase computational complexity and may not suit delay-sensitive practical applications such as real-time drone video transmission. If future motion can be predicted from external metadata, encoding can be optimized with lower complexity. In this study, a mathematical model for predicting motion vectors in drone video using only flight parameters is proposed. A remote-controlled drone with a fixed downward-facing camera recorded 4K video at 50 fps during autonomous flights over a marked terrain. Four flight parameters were varied independently, altitude, horizontal speed, vertical speed, and rotational rate. OpenCV was used to detect ground markers and compute motion vectors for temporal distances of 5 and 25 frames. Polynomial surface fitting was applied to derive motion models for translational, rotational, and elevational motion, which were later combined. The model was validated using complex motion scenarios (e.g., circular, ramp, helix), yielding worst-case prediction errors of approximately −1 ± 3 and −6 ± 14 pixels at 5 and 25 frames, respectively. The results suggest that flight-aware modeling enables accurate and low-complexity motion vector prediction for drone video coding. Full article

In automated orchard operations, the straight-line locomotion stability of ground-based weeding robots is critical for ensuring path coverage efficiency and operational reliability. To address the response lag and high-frequency oscillations often observed in conventional PID and fixed-lookahead Pure Pursuit controllers, this study proposes an adaptive lookahead Pure Pursuit method incorporating angular velocity feedback. By dynamically adjusting the lookahead distance according to real-time attitude changes, the method enhances coordination between path curvature and robot stability. To enable systematic evaluation, three time-series-based metrics are introduced: mean absolute yaw error (MAYE), peak-to-peak fluctuation amplitude, and the standard deviation of angular velocity, with overshoot occurrences included as an additional indicator. Field experiments demonstrate that the proposed method outperforms baseline algorithms, achieving lower yaw errors (0.61–0.66°), reduced maximum deviation (≤3.7°), and smaller steady-state variance (<0.44°²), thereby suppressing high-frequency jitter and improving turning convergence. Under typical working conditions, the method achieved a mean yaw deviation of 0.6602°, a fluctuation of 5.59°, an angular velocity standard deviation of 10.79°/s, and 155 overshoot instances. The yaw angle remained concentrated around the target orientation, while angular velocity responses stayed stable without loss-of-control events, indicating a favorable balance between responsiveness and smoothness. Overall, the study validates the robustness and adaptability of the proposed strategy in complex orchard scenarios and establishes a reusable evaluation framework, offering theoretical insights and practical guidance for intelligent agricultural machinery optimization. Full article

Hybrid underwater gliders (HUGs) combine buoyancy-driven gliding with propeller-assisted propulsion, offering extended endurance and enhanced mobility for complex underwater missions. However, precise depth control remains challenging due to system uncertainties, environmental disturbances, and inadequate adaptability of conventional control methods. This study proposes a novel optimized line-of-sight active disturbance rejection control (OLOS-ADRC) strategy for HUG depth tracking in the vertical plane. First, an Optimized Line-of-Sight (OLOS) guidance dynamically adjusts the look-ahead distance based on real-time cross-track error and velocity, mitigating error accumulation during path following. Second, a Tangent Sigmoid-based Tracking Differentiator (TSTD) enhances the disturbance estimation capability of the Extended State Observer (ESO) within the Active Disturbance Rejection Control (ADRC) framework, improving robustness against unmodeled dynamics and ocean currents. As a critical step before costly sea trials, this study establishes a high-fidelity simulation environment to validate the proposed method. The comparative experiments under gliding and hybrid propulsion modes demonstrated that OLOS-ADRC has significant advantages: the root mean square error (RMSE) for depth tracking was reduced by 83% compared to traditional ADRC, the root mean square error for pitch angle was decreased by 32%, and the stabilization time was shortened by 14%. This method effectively handles ocean current interference through real-time disturbance compensation, providing a reliable solution for high-precision HUG motion control. The simulation results provide a convincing foundation for future field validation in oceanic environments. Despite these improvements, the study is limited to vertical plane control and simulations; future work will involve full ocean trials and 3D path tracking. Full article

Weed infestation contributes significantly to global agricultural yield loss and increases the reliance on herbicides, raising both economic and environmental concerns. Effective weed detection in agriculture requires high accuracy and architectural efficiency. This is particularly important under challenging field conditions, including densely clustered targets, small weed instances, and low visual contrast between vegetation and soil. In this study, we propose GTDR-YOLOv12, an improved object detection framework based on YOLOv12, tailored for real-time weed identification in complex agricultural environments. The model is evaluated on the publicly available Weeds Detection dataset, which contains a wide range of weed species and challenging visual scenarios. To achieve better accuracy and efficiency, GTDR-YOLOv12 introduces several targeted structural enhancements. The backbone incorporates GDR-Conv, which integrates Ghost convolution and Dynamic ReLU (DyReLU) to improve early-stage feature representation while reducing redundancy. The GTDR-C3 module combines GDR-Conv with Task-Dependent Attention Mechanisms (TDAMs), allowing the network to adaptively refine spatial features critical for accurate weed identification and localization. In addition, the Lookahead optimizer is employed during training to improve convergence efficiency and reduce computational overhead, thereby contributing to the model’s lightweight design. GTDR-YOLOv12 outperforms several representative detectors, including YOLOv7, YOLOv9, YOLOv10, YOLOv11, YOLOv12, ATSS, RTMDet and Double-Head. Compared with YOLOv12, GTDR-YOLOv12 achieves notable improvements across multiple evaluation metrics. Precision increases from 85.0% to 88.0%, recall from 79.7% to 83.9%, and F1-score from 82.3% to 85.9%. In terms of detection accuracy, mAP:0.5 improves from 87.0% to 90.0%, while mAP:0.5:0.95 rises from 58.0% to 63.8%. Furthermore, the model reduces computational complexity. GFLOPs drop from 5.8 to 4.8, and the number of parameters is reduced from 2.51 M to 2.23 M. These reductions reflect a more efficient network design that not only lowers model complexity but also enhances detection performance. With a throughput of 58 FPS on the NVIDIA Jetson AGX Xavier, GTDR-YOLOv12 proves both resource-efficient and deployable for practical, real-time weeding tasks in agricultural settings. Full article

This paper proposes a speed-adaptive autonomous driving path-tracking framework based on the soft actor–critic (SAC) and pure pursuit (PP) methods, named the SACPP controller. The framework first analyzes the obstacles around the vehicle and plans an obstacle-free reference path with the minimum curvature using the hybrid A* algorithm. Next, based on the generated reference path, the current state of the vehicle, and the vehicle motor energy efficiency diagram, the optimal speed is calculated in real time, and the vehicle dynamics preview point at the future moment—specifically, the look-ahead distance—is predicted. This process relies on the learning of the SAC network structure. Finally, PP is used to generate the front wheel angle control value by combining the current speed and the predicted preview point. In the second layer, we carefully designed the evaluation function in the tracking process based on the uncertainties and performance requirements that may occur during vehicle driving. This design ensures that the autonomous vehicle can not only quickly and accurately track the path, but also effectively avoid surrounding obstacles, while keeping the motor running in the high-efficiency range, thereby reducing energy loss. In addition, since the entire framework uses a lightweight network structure and a geometry-based method to generate the front wheel angle, the computational load is significantly reduced, and computing resources are saved. The actual running results on the i7 CPU show that the control cycle of the control framework exceeds 100 Hz. Full article

With the exponential growth of engineering monitoring data, data-driven neural networks have gained widespread application in predicting retaining structure deformation in foundation pit engineering. However, existing models often overlook the spatial deflection correlations among monitoring points. Therefore, this study proposes a novel deep learning framework, CGCA (Convolutional Gated Recurrent Unit with Cross-Attention), which integrates ConvGRU and cross-attention mechanisms. The model achieves spatio-temporal feature extraction and deformation prediction via an encoder–decoder architecture. Specifically, the convolutional structure captures spatial dependencies between monitoring points, while the recurrent unit extracts time-series characteristics of deformation. A cross-attention mechanism is introduced to dynamically weight the interactions between spatial and temporal data. Additionally, the model incorporates multi-dimensional inputs, including full-depth inclinometer measurements, construction parameters, and tube burial depths. The optimization strategy combines AdamW and Lookahead to enhance training stability and generalization capability in geotechnical engineering scenarios. Case studies of foundation pit engineering demonstrate that the CGCA model exhibits superior performance and robust generalization capabilities. When validated against standard section (CX1) and complex working condition (CX2) datasets involving adjacent bridge structures, the model achieves determination coefficients (R²) of 0.996 and 0.994, respectively. The root mean square error (RMSE) remains below 0.44 mm, while the mean absolute error (MAE) is less than 0.36 mm. Comparative experiments confirm the effectiveness of the proposed model architecture and the optimization strategy. This framework offers an efficient and reliable technical solution for deformation early warning and dynamic decision-making in foundation pit engineering. Full article

Realizing path tracking is crucial for improving the accuracy and efficiency of unmanned vehicle operations. In this paper, a path tracking hierarchical control method based on DDPG-PP is proposed to improve the path tracking accuracy of tracked unmanned vehicles. Constrained by the objective of minimizing path tracking error, with the upper controller, we adopted the DDPG method to construct an adaptive look-ahead distance optimizer in which the look-ahead distance was dynamically adjusted in real-time using a reinforcement learning strategy. Meanwhile, reinforcement learning training was carried out with randomly generated paths to improve the model’s generalization ability. Based on the optimal look-ahead distance output from the upper layer, the lower layer realizes precise closed-loop control of torque, required for steering, based on the PP method. Simulation results show that the path tracking accuracy of the proposed method is better than that of the LQR and PP methods. The proposed method reduces the average tracking error by 94.0% and 79.2% and the average heading error by 80.4% and 65.0% under complex paths compared to the LQR and PP methods, respectively. Full article

Wheel loaders in mines and construction sites repeatedly load soil from a pile to load receivers. Automating this task presents a challenging planning problem since each loading’s performance depends on the pile state, which depends on previous loadings. We investigate an end-to-end optimization approach considering future loading outcomes and transportation costs between the pile and load receivers. To predict the evolution of the pile state and the loading performance, we use world models that leverage deep neural networks trained on numerous simulated loading cycles. A look-ahead tree search optimizes the sequence of loading actions by evaluating the performance of thousands of action candidates, which expand into subsequent action candidates under the predicted pile states recursively. Test results demonstrate that, over a horizon of 15 sequential loadings, the look-ahead tree search is 6% more efficient than a greedy strategy, which always selects the action that maximizes the current single loading performance, and 14% more efficient than using a fixed loading controller optimized for the nominal case. Full article

As a core component of agricultural machinery autonomous navigation, path tracking control holds significant research value. The pure pursuit algorithm has become a prevalent method for agricultural vehicle navigation due to its effectiveness at low speeds, yet its performance critically depends on the selection of the look-ahead distance. The conventional approaches require extensive parameter tuning due to the complex influencing factors, while fixed look-ahead distances struggle to balance the tracking accuracy and adaptability. Considerable effort is required to fine-tune the system to achieve optimal performance, which directly affects the accuracy of the path tracking and the results in the cumbersome task of selecting an appropriate goal point for the tracking path. To address these challenges, this paper introduces a pure pursuit algorithm for high-clearance sprayers in agricultural machinery, utilizing a self-adjusting look-ahead distance. By developing a kinematic model of the pure pursuit algorithm for agricultural machinery, an evaluation function is then employed to estimate the pose of the machinery and identify the corresponding optimal look-ahead distance within the designated area. This is done based on the principle of minimizing the overall error, enabling the dynamic and adaptive optimization of the look-ahead distance within the pure pursuit algorithm. Finally, this algorithm was verified in simulations and bumpy field tests under various different conditions, with the average value of the lateral error reduced by more than 0.06 m and the tuning steps also significantly reduced compared to the fixed look-ahead distance in field tests. The tracking accuracy has been improved and the applicability of the algorithm for rapid deployment has been enhanced. Full article

Look-ahead electromagnetic (EM) logging-while-drilling (LWD) plays an indispensable role in the prediction of deep and ultra-deep reservoirs. Traditional electromagnetic logging-while-drilling (EMLWD) and ultra-deep EMLWD technologies exhibit certain limitations in the real-time detection of ahead-of-bit formations, making it challenging to meet precision drilling requirements under complex well conditions, with the development of petroleum and gas geology and exploration progress I n the direction of deep, ultra-deep, and complex reservoirs. As a new LWD technology, look-ahead EMLWD enables real-time identification of formation structures, fluid distributions, and interface positions ahead of the drill bit during the drilling process by leveraging the propagation characteristics of EM. This capability provides critical decision-making support for wellbore trajectory optimization, drilling risk assessment, and reservoir evaluation. Therefore, this paper conducts research on theoretical methodologies for look-ahead EMLWD. Leveraging the Born geometric factor theory, we derive the expression for the 3D geometric factor spatial signal and analyze the sensitivity of each component related to look-ahead. Building on this foundation, we establish the sensitivity expression for look-ahead operations and investigate the impact of various antenna configurations on its signal. The results indicate that the coaxial component (gzz) and coplanar components (gxx and gyy) are the primary contributors to look-ahead EMLWD. As frequency decreases and spacing increases, the sensitive region for look-ahead expands. Moreover, look-ahead detection sensitivity becomes increasingly concentrated in front of the drill bit, while the signal at the opposite end is attenuated by incorporating additional coils. Under identical formation conditions, compared with a single-transmitter single-receiver system, a single-transmitter double-receiver coil system exhibits a significantly stronger signal amplitude and more pronounced changes at the formation boundary. Additionally, this configuration enhances sensitivity and extends the sensitive distance in response to variations in formation resistivity. Full article

This paper presents a method for designing a static output feedback integrated path tracking controller for autonomous vehicles. For path tracking, state–space model-based control methods, such as linear quadratic regulator, H_∞ control, sliding mode control, and model predictive control, have been selected as controller design methodologies. However, these methods adopt full-state feedback. Among the state variables, the lateral velocity, or the side-slip angle, is hard to measure in real vehicles. To cope with this problem, it is desirable to use a state estimator or static output feedback (SOF) control. In this paper, an SOF control is selected as the controller structure. To design the SOF controller, a linear quadratic optimal control and sliding mode control are adopted as controller design methodologies. Front wheel steering (FWS), rear wheel steering (RWS), four-wheel steering (4WS), four-wheel independent braking (4WIB), and driving (4WID) are adopted as actuators for path tracking and integrated as several actuator configurations. For better performance, a lookahead or preview function is introduced into the state–space model built for path tracking. To verify the performance of the SOF path tracking controller, simulations are conducted on vehicle simulation software. From the simulation results, it is shown that the SOF path tracking controller presented in this paper is effective for path tracking with limited sensor outputs. Full article

Based on the characteristics of small agricultural machinery in terms of flexibility and high efficiency when operating in small plots of hilly and mountainous areas, as well as the demand for improving the automation and intelligence levels of agricultural machinery, this paper conducted research on the path tracking control of the automatic navigation operation of a crawler sprayer. Based on the principles of the kinematic model and the position prediction model of the agricultural machinery chassis, a pure pursuit controller based on adaptive look-ahead distance was designed for the tracked motion chassis. Using a lightweight crawler sprayer as the research platform, integrating onboard industrial control computers, sensors, communication modules, and other hardware, an automatic navigation operation system was constructed, achieving precise control of the crawler sprayer during the path tracking process. Simulation test results show that the path tracking control method based on adaptive look-ahead distance has the characteristics of smooth control and small steady-state error. Field tests indicate that the crawler sprayer exhibits small deviations during path tracking, with an average absolute error of 2.15 cm and a maximum deviation of 4.08 cm when operating at a speed of 0.7 m/s. In the line-following test, with initial position deviations of 0.5 m, 1.0 m, and 1.5 m, the line-following times were 7.45 s, 11.91 s, and 13.66 s, respectively, and the line-following distances were 5.21 m, 8.34 m, and 9.56 m, respectively. The maximum overshoot values were 6.4%, 10.5%, and 12.6%, respectively. The autonomous navigation experiments showed a maximum deviation of 5.78 cm and a mean absolute error of 2.69 cm. The proportion of path deviations within ±5 cm and ±10 cm was 97.32% and 100%, respectively, confirming the feasibility of the proposed path tracking control method. This significantly enhanced the path tracking performance of the crawler sprayer while meeting the requirements for autonomous plant protection spraying operations. Full article

This study introduces a motion-sickness-reducing control strategy aimed at enhancing ride comfort in Electric Autonomous Vehicles (EAVs). For lateral control, the forward look-ahead distance was adaptively adjusted based on the Motion Sickness Dose Value (MSDV) analysis from ISO 2631-1, effectively mitigating lateral acceleration and its motion-sickness-related frequency components, leading to a reduced MSDV. For longitudinal control, Linear Quadratic Regulator (LQR) optimal control was applied to minimize acceleration, complemented by a band-stop filter specifically designed to attenuate motion-sickness-inducing frequencies in the acceleration input. The bandwidth of the band-stop filter used in this study was designed based on the motion-sickness frequency weighting specified in ISO 2631-1. The simulation results of the proposed control indicate a significant reduction in MSDV, decreasing from 16.3 to 10.46, achieving up to a 35.8% improvement compared to comparative control methods. While the average lateral position error was slightly higher than that of the comparative controller, the vehicle consistently maintained lane adherence throughout path-following tasks. These findings underscore the potential of the proposed method to simultaneously mitigate motion sickness and achieve a robust path-following performance in autonomous vehicles. Full article