A multi-objective active front steering (AFS) control system considering the road adhesion constraint on vehicle stability is developed using the sliding mode control (SMC) method. First, an identification function combined with the relationship between the yaw rate and the steering angle is developed to determine whether the tyre state is linear or nonlinear. On this basis, an intervention criterion for the AFS system is proposed to improve vehicle handling and stability in emergent conditions. A sideslip angle stability domain enveloped by the upper, lower, left, and right boundaries, as well as the constraint of road adhesion coefficient, is constructed based on the phase-plane method. A dynamic weighting coefficient to coordinate the control of yaw rate and sideslip angle, and a control strategy that considers changing control objectives based on the desired yaw rate, the desired sideslip angle, and their proportional weights, are proposed for the SMC controller. Because road adhesion has a significant effect on vehicle stability and to meet the control algorithm's requirement of real-time access to vehicle states, a unscented Kalman filter-based state observer is proposed to estimate the adhesion coefficient and the required states. Finally, simulations are performed using high and low road adhesion conditions in a Matlab/Simulink environment, and the results show that the proposed AFS control system promptly intervenes according to the intervention criterion, effectively improving vehicle handling and stability.
15 November 2022 in Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering
When driving on a curved lane with varying curvature, human drivers usually make a small range of longitudinal speed adjustments to maintain a yaw response that feels more stable. Taking this type of response as a reference, intelligent vehicles must also make adjustments to their longitudinal speed within a small range during the path tracking process with curvature changes. The current variable curvature path tracking algorithm using model predictive control (MPC) basically assumes that the vehicle moves at a constant speed, which does not match the small-range adjustment of the longitudinal speed and thereby affects the accuracy of path tracking. In this paper, considering the small range of speed variation in the path tracking process, the path and speed decoupling control in Frenet coordinates are used to replace the longitudinal-lateral-yaw complex coupling dynamics control. Meanwhile, considering the problem that the steady-state error of the MPC controller caused by curvature variation in the path tracking process cannot be eliminated, the adaptive weight control (AWC) and adaptive feedforward (AFF) models based on BP neural network (BPNN) data learning are designed to dynamically adjust the lateral error weight and feedforward factors of the MPC controller. As a result, a more accurate path tracking effect is achieved. Simulation results in the joint CarSim-Simulink environment show that the proposed algorithm significantly improves the adaptive capability of the linear MPC controller in response to time-varying conditions and has a higher tracking accuracy.
13 September 2023 in Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering
A well-controlled active suspension system has the potential to provide better ride comfort. Benefiting from its powerful feature extraction and nonlinear generalization capabilities, the deep reinforcement learning (DRL), such as deep deterministic policy gradient (DDPG), has shown great potential to make decisions adaptively and intelligently in the control of active suspension system. However, the DDPG is troubled by the problem of low training efficiency due to the high proportion of illegal strategies. This paper proposed a novel DDPG controller for a nonlinear uncertain active suspension system by combining DRL with expert demonstrations. Specifically, the improved training method integrated with both a pre-training mechanism based on PID expert samples and an adaptive experience replay mechanism, is put forward for the DDPG to achieve both the goals of imitating the expert and improving the training efficiency. Moreover, considering the ride comfort and the state constraints as targets, a mixed reward function is designed to guide RL agents for learning effective actions. It is shown that the proposed training methods effectively accelerate the convergence of the DDPG. Furthermore, the comparison experiments demonstrate that the proposed controller provides great vibration attenuation, and has better adaptiveness to various working conditions and parametric uncertainty.
10 June 2021 in Industrial Robot: the international journal of robotics research and application
Purpose Trajectory tracking error of robotic manipulator has limited its applications in trajectory tracking control systems. This paper aims to improve the trajectory tracking accuracy of robotic manipulator, so a linear-extended-state-observer (LESO)-based prescribed performance controller is proposed. Design/methodology/approach A prescribed performance function with the convergence rate, maximum overshoot and steady-state error is derived for the output error transformation, whose stability can guarantee trajectory tracking accuracy of the original robotic system. A LESO is designed to estimate and eliminate the total disturbance, which neither requires a detailed system model nor a heavy computation load. The stability of the system is proved via the Lyapunov theory. Findings Comparative experimental results show that the proposed controller can achieve better trajectory tracking accuracy than proportional-integral-differential control and linear active disturbance rejection control. Originality/value In the LESO-based prescribed performance control (PPC), the LESO was incorporated into the PPC design, it solved the problem of stabilizing the complex transformed system and avoided the costly offline identification of dynamic model and estimated and eliminated the total disturbance in real-time with light computational burden. LESO-based PPC further improved control accuracy on the basis of linear-active-disturbance-rejection-control. The new proposed method can reduce the trajectory tracking error of the robotic manipulators effectively on the basis of simplicity and stability.
30 August 2024 in IEEE Transactions on Vehicular Technology
Driver-automation cooperative driving has recently been proved to be a promising approach to enhance driving safety in common or even relatively complex traffic scenarios. However, the potential of this cooperative driving system to enhance driving safety under extreme conditions remains underexplored. This paper presents a game-theoretic shared steering control (SSC) strategy, which takes nonlinear vehicle dynamics as well as state and input constraints into consideration, for driver-automation cooperative collision avoidance under extreme conditions. Firstly, a novel Takagi-Sugeno (T-S) fuzzy Koopman operator approach for vehicle system modeling is introduced. This approach ensures the accuracy of the Koopman-based vehicle model in the presence of changing external conditions, such as road adhesion coefficient and vertical load. The resulting vehicle model maintains a simple linear structure while preserving system nonlinearity. Then, based on the T-S fuzzy Koopman vehicle model, we design a game-theoretic SSC strategy that considers state and input constraints. The state constraints are further approximated through artificial potential field functions and second-order Taylor series expansions, enabling the derivation of analytical Nash equilibrium solution using the distributed model predictive control approach. Finally, the effectiveness of the proposed SSC strategy is validated through both numerical and driver-in-the-loop (DiL) experiments. The results indicate that the proposed SSC strategy significantly improves the vehicle's collision-avoidance performance while reducing the driver's workloads.
1 January 2019 in International Journal of Vehicle Design
To improve a vehicle's anti-roll performance while consuming less energy, an active/passive dual-mode switched control method based on a hydraulically interconnected suspension (HIS) is proposed. This method has two key components as follows: modelling the mechanical-liquid-gas coupled dynamic system consisting of hydraulic cylinders, a hydraulic pump and accumulators; and robustly controlling a time-delayed system with uncertain model and lag time. In this study, the nonlinear characteristics of the underlying actuators were considered, and a four degrees-of-freedom (4DOF) roll-plane half-car model was employed. The backstepping algorithm was adopted to maintain the body in the desired posture. Subsequently, a modified Smith predictor and a Taylor expansion of lateral acceleration were applied simultaneously to compensate for the time delay. Finally, simulations with uncertain parameters and lag time were performed. The results indicate that the proposed method is robust in the case of an uncertain control system and can effectively improve the vehicle's anti-roll performance.