Research On Path Tracking Control Strategy Of Four-wheel-steering And Four-wheel-drive Vehicle

With the development of unmanned ground vehicles in the past few years,the demand for high accuracy in navigation is ever increasing.The four-wheel-steering and four-wheel-drive(4WS4WD)vehicle platform has great potential to solve existing problems,such as path tracking inaccuracy due to wheel slip and various forces generated on complex terrain.Although the configuration of 8 control inputs can enhance the vehicle's maneuverability,it turns the vehicle into an over-actuated,highly coupled and nonlinear system,which poses a challenge for the control system design.The work in this thesis is to develop a control strategy for 4WS4WD vehicle precise path tracking.Firstly,a novel control method of the force-controlled 4WS4WD vehicle is proposed in which the vehicle is controlled by the tire longitudinal force,Accordingly,a kinematic model and a modular dynamic model composed of vehicle body dynamic model,driving unit dynamic model and tire model are developed to facilitate independent steering and PID based force control at four individual wheels.The effects of kinematic model and dynamic model are compared in the simulation,the effectiveness of the proposed dynamic model is verified,which lay a foundation for the follow-up research of 4WS4WD vehicle path tracking control strategy.In order to describe the 4WS4WD vehicle path tracking,the offset models are established based on the kinematic model and dynamic model respectively.The MPC-LCA control strategy is proposed based on the offset models,which separates the nonliner control problem into an upper layer linear tracking problem and a lower layer control allocation problem.The upper layer controller is used to calculate the steering angles and virtual accelerations based on a proposed offline simplified model predictive control algorithm.By using the results from the upper layer,the lower layer controller can optimize the drive forces based on the active set method.Simulation results are provided to demonstrate the effectiveness and applicability of the proposed strategy.In order to avoid the sub-optimal solution of the control inputs by linear control allocation,the nonlinear control allocation is proposed to replace the linear control allocation in the lower layer controller.Two proposed nonlinear control allocation algorithms are based on the sequential quadrantic programming(SQP)and paticle swarm optimization(PSO),respectively.The improvement is to optimize the steering angles and drive forces synchronously based on the dynamic model instead of calculating the steering angles relying on kinematic model,which leads to better tracking accuracy of 4WS4WD vehicle.The parameter setting law of two algorithms are obtained by the case simulation.The result of the algorithms implemented in 4WS4WD vehicle path tracking is analysed,which indicates a better path tracking effect.To solve the outer disturbance issue during path tracking,a multi-disturbance robust model predictive comtrol algorithm is proposed by considering the constraints of H2 and H? performance indicators,which can improve the robustness of the upper layer controller.The multi disturbance model is obtained by analyzing the uncertainty and disturbance in the offset model.The proposed robust predictive control algorithm can adaptively fix the tracking errors with the H2 constraint and restrain different kind of disturbance with the H? constraint.Combining the robust predictive control and SQP based nonlinear control allocation,a complete 4WS4WD vehicle path tracking control strategy is obtained,named as RMPC-SQPCA strategy.The proposed strategy guarantees the precise tracking of the vehicle as well as the robustness in the simulation.Finally,the experimental prototype of the FC-4WS4WD vehicle is established.By running the RMPC-SQPCA strategy on the experimental prototype,the validity of the proposed strategy is verified.