Game Based Coordinated And Driving Fault-Tolerant Control Of Independently Driving Electric Vehicle

Since traditional fuel vehicles have caused serious energy crisis and environmental pollution,electric vehicles have the incomparable advantages in terms of environmental protection and energy saving.The motion state of each wheel in Four Wheel Independently Driving Electric Vehicles(FWID-EV)is independently controllable and it is easy to realize complex control,so it gradually become a hot topic in today's research.Since there is no rigid mechanical connection between the wheels of this kind vehicle,and its driving/braking torque is independently controllable,so it has more controllable degrees of freedom,which brings great challenge to the stability control of the vehicle.At the same time,with the increase of the number of actuators,the probability of actuator failure has also increased significantly,so its safety problem has become a hot spot in the automotive field.This paper focuses on the stability coordination control and driving failure of electric vehicles,and using game theory and model predictive control theory to conduct a series of researches on lateral stability control and driving fault-tolerant control of electric vehicles.The main research work includes the following points:(1)The construction of system dynamic model for FWID-EV.The electric vehicle was rebuilt based on traditional car(B-Class Car)in CarSim.The vehicle model of the electric vehicle with X-by-driving and steering was established by the combination of CarSim and MATLAB/Simulink software,and the model was verified by Simulation;(2)Non-cooperative game theory based lateral stability control research.Based on the overdriving structure of FWID-EV,a hierarchical control structure based on non-cooperative game is proposed,which is divided into upper and lower layers.The upper controller designs the stability control strategy based on Nash and Stackelberg game theory,respectively,in which the corresponding control behavior can be obtained based on the control strategy.The lower controller optimizes the virtual control values(additional yaw moment)of the upper controller by using the Lagrangian multiplier optimization algorithm;(3)Cooperative game theory based lateral stability control research.A similar two-layer hierarchical control structure is adopted,in which the upper controller designs the stability control strategy based on the cooperative game theory,and the corresponding virtual control actions are obtained based on the strategy.A sequential quadratic programming(SQP)optimization allocation algorithm is used in the lower layer controller to distribute the virtual control actions(additional yaw moment)to each wheel's actuator;(4)Model prediction(MPC)based fault-tolerant control research.Firstly,a three-degree-of-freedom vehicle dynamic control model including lateral,longitudinal and yaw motion for MPC is established.Then,the cost function is designed to reduce the vehicle's sideslip angle and yaw rate deviation,and the constraint is established for the controller output.At the same time,the vehicle control structure is reconstructed after the vehicle fails,and the reconstructed prediction model is used to estimate the future state and output of the vehicle.Finally,the quadratic programming optimization algorithm is used to solve the multi-constraint optimization problem in the rolling time domain.Finally,the effectiveness of designed control strategy is verified by the joint simulation of CarSim and MATLAB/Simulink.The research results show that the non-cooperative and cooperative game based vehicle stability control strategy can maintain the vehicle in a relatively stable state under extreme conditions.Among them,the cooperative game control reduces the peak value of the vehicle's sideslip angle by 63.1% compared with the classic LQR controller in the double-lane change test condition with high speed and low road adhesion coefficient.At the same time,the MPC fault-tolerant controller can keep the vehicle stay in a safe state through the fault-tolerant control of functional redundant healthy steering and drive/brake system under the condition that the left front wheel fails.