Path Tracking Method Research Articles

A knowledge-guided reinforcement learning method for lateral path tracking

Lateral Control algorithms in autonomous vehicles often necessitates an online fine-tuning procedure in the real world. While reinforcement learning (RL) enables vehicles to learn and improve the lateral control performance through repeated trial and error interactions with a dynamic environment, applying RL directly to safety-critical applications in real physical world is challenging because ensuring safety during the learning process remains difficult. To enable safe learning, a promising direction is to make use of previously gathered offline data, which is frequently accessible in engineering applications. In this context, this paper presents a set of knowledge-guided RL algorithms that can not only fully leverage the prior collected offline data without the need of a physics-based simulator, but also allow further online policy improvement in a smooth, safe and efficient manner. To evaluate the effectiveness of the proposed algorithms on a real controller, a hardware-in-the-loop and a miniature vehicle platform are built. Compared with the vanilla RL, behavior cloning and the existing controller, the proposed algorithms realize a closed-loop solution for lateral control problems from offline training to online fine-tuning, making it attractive for future similar RL-based controller to build upon.

Path tracking with switchable stability constraints for vehicle collision avoidance based on extended stability region

There is a certain conflict between vehicle stability and the performance of path-tracking under emergency conditions. In this paper, to make full use of the lateral capacity of vehicles, a path tracking method with switchable stability constraints is proposed. First, using the Lyapunov energy equation and Hurwitz’s theorem, the lateral stability region (SR) and the critical steering angle of steady steering at different longitudinal speeds are calculated. Based on this, the extended stability region (ESR) and the extended critical steering angle of the vehicle under the control of an additional yaw moment are obtained. Furthermore, the driving area during collision avoidance is divided into unsafe and safe areas, and based on the model predictive control strategy, a path-tracking controller with switchable stability constraints is established. In the unsafe area, the vehicle states are allowed to break through the SR temporarily, and to ensure path-tracking accuracy, the ESR is taken as the stability constraint. In the safe area, the stability constraint is switched to the SR, and to make the vehicle states quickly return to SR, a regain stability controller is established. Finally, a Simulink-CarSim simulation and a hardware-in-the-loop (HIL) experiment are conducted. The simulation results show that a balance between tracking effect and vehicle stability can be effectively achieved through the proposed path-tracking system under the investigated emergency conditions. The real-time performance of the proposed control strategy is validated through the HIL experiment.

Design of a Path-Following Controller for Autonomous Vehicles Using an Optimized Deep Deterministic Policy Gradient Method

The need for a safe and reliable transportation system has made the advancement of autonomous vehicles (Avs) increasingly significant. To achieve Level 5 autonomy, as defined by the Society of Automotive Engineers, AVs must be capable of navigating complex and unconventional traffic environments. Path-following is a crucial task in autonomous driving, requiring precise and safe navigation along a defined path. Traditional path-tracking methods often rely on parameter tuning or rule-based approaches, which may not be suitable for dynamic and complex environments. Reinforcement learning has emerged as a powerful technique for developing effective control strategies through agent-environment interactions. This study investigates the efficiency of an optimized Deep Deterministic Policy Gradient (DDPG) method for controlling acceleration and steering in the path-following of autonomous vehicles. The algorithm demonstrates rapid convergence, enabling stable and efficient path tracking. Additionally, the trained agent achieves smooth control without extreme actions. The performance of the optimized DDPG is compared with the standard DDPG algorithm, with results confirming the improved efficiency of the optimized approach. This advancement could significantly contribute to the development of autonomous driving technology.

A Velocity-Adaptive MPC-Based Path Tracking Method for Heavy-Duty Forklift AGVs

A Velocity-Adaptive MPC-Based Path Tracking Method for Heavy-Duty Forklift AGVs

Improved Pure Pursuit Algorithm Based Path Tracking Method for Autonomous Vehicle

Improved Pure Pursuit Algorithm Based Path Tracking Method for Autonomous Vehicle

A Novel Robust H∞ Control Approach Based on Vehicle Lateral Dynamics for Practical Path Tracking Applications

This paper proposes a robust lateral control scheme for the path tracking of autonomous vehicles. Considering the discrepancies between the model parameters and the actual values of the vehicle and the fluctuation of parameters during driving, the norm-bounded uncertainty is utilized to deal with the uncertainty of model parameters. Because some state variables in the model are difficult to measure, an H∞ observer is designed to estimate state variables and provide accurate state information to improve the robustness of path tracking. An H∞ state feedback controller is proposed to suppress system nonlinearity and uncertainty and produce the desired steering wheel angle to solve the path tracking problem. A feedforward control is designed to deal with road curvature and further reduce tracking errors. In summary, a path tracking method with H∞ performance is established based on the linear matrix inequality (LMI) technique, and the gains in observer and controller can be obtained directly. The hardware-in-the-loop (HIL) test is built to validate the real-time processing performance of the proposed method to ensure excellent practical application potential, and the effectiveness of the proposed control method is validated through the utilization of urban road and highway scenes. The experimental results indicate that the suggested control approach can track the desired trajectory more precisely compared with the model predictive control (MPC) method and make tracking errors within a small range in both urban and highway scenarios.

Design and Testing of an Autonomous Navigation Unmanned Surface Vehicle for Buoy Inspection

In response to the inefficiencies and high costs associated with manual buoy inspection, this paper presents the design and testing of an Autonomous Navigation Unmanned Surface Vehicle (USV) tailored for this purpose. The research is structured into three main components: Firstly, the hardware framework and communication system of the USV are detailed, incorporating the Robot Operating System (ROS) and additional nodes to meet practical requirements. Furthermore, a buoy tracking system utilizing the Kernelized Correlation Filter (KCF) algorithm is introduced. Secondly, buoy image training is conducted using the YOLOv7 object detection algorithm, establishing a robust model for accurate buoy state recognition. Finally, an improved Line-of-Sight (LOS) method for USV path tracking, assuming the presence of an attraction potential field around the inspected buoy, is proposed to enable a comprehensive 360-degree inspection. Experimental testing includes validation of buoy image target tracking and detection, assessment of USV autonomous navigation and obstacle avoidance capabilities, and evaluation of the enhanced LOS path tracking algorithm. The results demonstrate the USV’s efficacy in conducting practical buoy inspection missions. This research contributes insights and advancements to the fields of maritime patrol and routine buoy inspections.

A tube-based model predictive control method for intelligent vehicles path tracking

A tube-based model predictive control method for intelligent vehicles path tracking

Neuroadaptive Fault-Tolerant Control Embedded With Diversified Activating Functions With Application to Auto-Driving Vehicles Under Fading Actuation.

This article presents a neuroadaptive fault-tolerant control method for path tracking of multiinput multioutput (MIMO) systems in the presence of modeling uncertainties and external disturbances. In dealing with modeling uncertainties, neural networks (NNs) with diversified activation/basis functions are considered, with which we establish a set of control algorithms that are robust against uncertainties, adaptive to unknown parameters, and tolerant to actuation faults. This is the first work that explicitly takes into account the neural weights uncertainties and activating function uncertainties in multiple layered neural networks in control design. In addition, we apply the developed control algorithms to unmanned ground vehicles (UGVs) with actuator failures. With the aid of Lyapunov stability theory, it is shown that the proposed control is able to drive the vehicle along the desired path with high precision and all the internal signals are uniformly ultimately bounded (UUB) and continuous. Both theoretical analysis and numerical simulation confirm the effectiveness of the designed strategy.

Trajectory Tracking Control of Variable Sweep Aircraft Based on Reinforcement Learning.

An incremental deep deterministic policy gradient (IDDPG) algorithm is devised for the trajectory tracking control of a four-wing variable sweep (FWVS) aircraft with uncertainty. The IDDPG algorithm employs the line-of-sight (LOS) method for path tracking, formulates a reward function based on position and attitude errors, and integrates long short-term memory (LSTM) units into IDDPG algorithm to enhance its adaptability to environmental changes during flight. Finally, environmental disturbance factors are introduced in simulation to validate the designed controller's ability to track climbing trajectories of morphing aircraft in the presence of uncertainty.

A model-free adaptive predictive path-tracking controller with PID terms for tractors

An efficient tractor path-tracking method can increase the operational accuracy, improve land utilisation, and more efficiently provide services for autonomous tractors and precision agriculture. In this study, a model-free adaptive predictive control-proportional-integral-derivative (MFAPC-PID) method was devised. Based on preview theory, a tracking system based on course deviation angle was designed. After the tracking system was linearised, an MFAPC tracking system was developed, and its inherent defects were analysed. Referring to the control structure of an incremental PID algorithm, the MFAPC was considered as an adaptive integral item, and the MFAPC-PID controller was obtained by adding adaptive proportional and differential terms. To verify the proposed controller, collaborative simulation and hardware-in-the-loop tests were performed. The controller was simulated and tested under different paths, road surfaces, and tractor models. Compared with other methods, the MFAPC-PID path-tracking method exhibits superior comprehensive performance, adaptability, universality, and robustness. Moreover, the MFAPC-PID method is insensitive to external interference and model changes of controlled objects.

A three level NPC inverter fault detection and identification using voltage vector residual evaluation based on current path tracking method

This research describes open circuit detection and fault location for active 3-level NPC (neutral point clamped) inverter switches during normal and connected to the neutral point. This work introduces the current path tracking method to correlate the operating mode of NPC inverter and switch status in abnormal and normal conditions. Furthermore, a hierarchal diagnosis approach is introduced for open circuit faults using output Voltage Vector Residual Evaluation (VVRE). The node potential and conduction path can be derived from the model to indicate the NPC inverter circuit faults. Two steps are involved in the diagnosis. First, OC fault can be determined by analyzing the difference between the reference voltage and actual voltage vector resulting from the specified current path utilization ratio. Therefore, using a current path state reconstruction method, we can distinguish faulty transistors with similar properties. The simulated results demonstrate the robustness and efficiency of the proposed approach.

A Fast Forward Modelling Method for Simulating Satellite Observations Using Observing Path Tracking

The higher the atmosphere is, the larger the deviations in atmospheric temperature and humidity are between the vertical column atmosphere above the cross-section of a satellite instrument and a ray’s trajectory from the cross-section to the satellite. In general, satellite instruments that observe using cross-orbit scanning result in the difference between the observed radiance and the simulations using this method becoming incrementally larger and larger as the cross-section moves to the edge of the satellite’s orbit. The deviations depend on the distance from the column to the ray trajectory and on the horizontal gradient of variables in the distance. In fact, the horizontal gradient of water vapour is larger than the gradient of temperature in clear scenarios, which could introduce an impact of temperature and water vapour on the simulated radiance of a satellite. In this study, a new method to simulate upgoing and downgoing radiation synchronously was developed, using the observing path tracking method. The conventional vertical initial atmospheric profile (Exp.1) and the profiles along the upgoing and downgoing rays of the satellite’s observation (Exp.2) were established, in order to simulate the observed radiance of MWHS-II of FY-3D using global numerical forecasts with resolutions of 15 km and 25 km. The results showed that, for channels in the oxygen and water vapour absorption line on the microwave spectrum, deviations of the two atmospheric profiles were larger at the scan edge (0.01 K) than those at the nadir (0.001 K), and were larger in the upper atmosphere than in the lower atmosphere. The deviation was usually negative in low-latitude regions and was positive in southern high-latitude regions. Such results were obtained in experiments using both the numerical forecast method with 15 km grids and the forecast method with 25 km grids. Deviations were analysed for representative channels at 118 GHz and 183 GHz. Then, the results indicated that bigger deviations between the two experiments were observed in the water vapour absorption line than in the oxygen absorption line in the microwave spectrum. In conclusion, this indicates that, because of the greater horizontal gradient of water vapour, the stronger localisation of water vapour makes the atmospheric profile along the satellite’s observing ray have more increments in the simulated radiance at the scan edge, compared to the atmospheric column profile.

Multi-AUV coverage path planning algorithm using side-scan sonar for maritime search

Coverage search is widely applied in the maritime military and civilian fields. To improve search efficiency and detection accuracy, we propose an effective and rapid multi-autonomous underwater vehicle (AUV) coverage path planning (ER-MCPP) algorithm using side-scan sonar (SSS). First, the given search region was optimized and modeled using grid decomposition. Next, a Voronoi diagram was used to divide and allocate the regions according to the initial position of the AUV. Subsequently, based on the idea of guidance, navigation, and control (GNC), a coverage search algorithm for a single AUV was proposed, including a path planning method combining improved lawn mower and outward-spiral methods, an integrated navigation method based on Kalman filtering, and a path tracking method considering ocean currents. In addition, we proposed a fault-tolerant control method. When an AUV fails unexpectedly, the other AUVs can accomplish a coverage search for the region through cooperation. Simulation experiments were performed using the MATLAB software. The results demonstrate that the proposed algorithm is valid. Compared with other methods, our algorithm has better robustness, finds targets more rapidly, and has a higher search efficiency. The practicality of the proposed algorithm is verified through field tests.

A Fast Predictive Control Method for Vehicle Path Tracking Based on a Recurrent Neural Network

A Fast Predictive Control Method for Vehicle Path Tracking Based on a Recurrent Neural Network

An improved model predictive control method for path tracking of autonomous vehicle considering longitudinal velocity

In order to increase the accuracy of the path tracking, an improved model predictive control (IMPC) is proposed for autonomous vehicle under road conditions of large curvature, which can enhance the performances of the driving stability and safety. The controller design is implemented in four steps. First, the curvature of road ahead is derived and applied to determine the longitudinal velocity. Thus, the longitudinal velocity is not assumed to be constant, which is the salient feature of the proposed control. Second, the kinematic model of vehicle is established by the Ackermann steering principle. Third, the predictive model is constructed by linearization and discretization of the kinematic model. Fourth, the longitudinal velocity and the front steering angle are imposed on hard constraints, and the constrained objective function is designed and composed of the position deviation and the control increment. Then, we can obtain the optimal results of the longitudinal velocity and the front steering angle. Furthermore, experiment and simulation on the path tracking of an autonomous vehicle are presented. The results show that the proposed control can realize excellent tracking performance under the road conditions of large curvature.

Path Tracking of Agricultural Vehicles Based on 4WIS–4WID Structure and Fuzzy Control

This paper proposes a path-tracking method for agricultural vehicles based on the 4WIS-4WID structure and fuzzy control to improve the operation performance of agricultural machinery in the greenhouse. The influential model of two critical parameters, α and R, on the position correction is obtained based on the relationship analysis between the traditional pure pursuit model and the vehicle structure. Based on aiming pursuit, the relationship equation between the position deviation, lateral deviation d, and heading deviation θ is established. A two-input and two-output fuzzy controller is designed, and the lateral deviation d and heading deviation θ are the input variables. Values of α and R are obtained after fuzzification, fuzzy inference, and defuzzification, and a validated MATLAB model is used to simulate different scenarios. Results of the tests show that the steady-state error of path tracking based on fuzzy control pursuit is between 35 and 51 mm, and the stability distance is between 1661 and 3052 mm for straight path tracking in four initial states. The rectangular corners have the highest inaccuracy. The results of fuzzy control pursuit show a significant improvement in path-tracking performance that can influence vehicle navigation capability in the greenhouse.

Construction path tracking and pose estimation of unmanned bulldozer

The vigorous development of automatic control technology and AI has generated considerable interest in driverless systems. Presently, research on driverless systems is mainly focused on the automotive field, but it also has immense scope in engineering machinery. In the field of unmanned bulldozers, path tracking and pose estimation are the foundation of construction quality and safety. This article focuses on this field. A construction path tracking method was designed, wherein motion control commands were generated according to the course angle deviation combined with verification of steering gain and saturation characteristics. The proposed method can track the construction path of an unmanned bulldozer with high accuracy. The improved ORB-SLAM3 pose estimation method fused with GNSS and the local map reloading pose estimation method are used to estimate the bulldozer pose to ensure construction safety. The two methods are superior to the general SLAM system in terms of precision and dynamic performance.

Research on AGV path tracking method based on global vision and reinforcement learning.

As the key to the movement of automated guided vehicle (AGV), the design of control algorithm directly affects whether AGV can follow the preset path. Aiming at the difficulty of AGV control, an AGV path tracking control method based on global vision and reinforcement learning is proposed. Firstly, the global view is obtained by the visual sensor, and the position information of obstacles and AGV is obtained by the target detection algorithm. Secondly, the path planning algorithm is used to obtain the driving path information which is used to establish a virtual environment. Thirdly, the position and pose of the physical AGV are introduced into the virtual environment by the visual sensor, and the virtual AGV is reset. Finally, the image obtained by virtual vehicle camera is input into the reinforcement learning model and the output action is sent to the physical AGV for execution. In the experimental part, this method can not only plan the driving path in different environments but also well control AGV to drive along the specified path, which proves that this method has strong robustness and feasibility.

Integrated control method for path tracking and lateral stability of distributed drive electric vehicles with extended Kalman filter–based tire cornering stiffness estimation

Aiming at the lack of adaptability of vehicle parameters under extreme conditions, this paper proposes an integrated control method for path tracking and lateral stability of distributed drive electric vehicles based on tire cornering stiffness adaptive model predictive control (MPC) scheme. The control method integrates active front steering and direct yaw control to improve the path tracking and lateral stability performance of distributed drive electric vehicles. Firstly, considering the influence of vertical load transfer, the tire cornering stiffness is estimated based on the extended Kalman filter (EKF) algorithm. Then, using this online updated tire cornering stiffness value, an adaptive MPC controller for path tracking and lateral motion stability of distributed drive electric vehicles is constructed. Meanwhile, a fuzzy sliding mode control (Fuzzy-SMC)–based longitudinal velocity controller is established to ensure the accuracy of velocity tracking. Also, according to the distributed driving characteristics of the controlled system, a tire torque distributor based on weighted pseudo-inverse (WPI) is designed with the minimum tire load rate as the optimization objective, where the road adhesion condition and the maximum output torque of the motor are considered as constraints. The simulation results show that the proposed integrated control method based on tire cornering stiffness adaptive model predictive control is robust and effective. Compared with constant cornering stiffness model predictive control–based control method, it can improve the vehicle path tracking accuracy and lateral motion stability under extreme conditions.