Vehicle Obstacle Avoidance Path Planning And Tracking Control Based On Adaptive Model Predictive Control

Nowadays,the traffic is deteriorating day by day,which leads to frequent traffic accidents.According to statistics,driver's misoperation is one of the main reasons for accidents.Intelligent vehicles can reduce human intervention,so it is of great significance to research intelligent vehicles for improving driving safety,vehicle stability and traffic efficiency.Firstly,to solve the issue of path tracking in complex and changeable environment,a path tracking control framework based on adaptive model predictive control(AMPC)is proposed.According to the recursive least square method,the cornering stiffness estimator and the road adhesion coefficient estimator are designed to obtain the tire parameters and road conditions.In addition,the parameter selection module is designed,which selects control parameters according to vehicle speed.Then,taking the three degrees of freedom vehicle dynamics model as the prediction model,a path tracking controller is designed with model predictive control(MPC)algorithm.To track the target path,the optimal front wheel angle is obtained by solving the optimization problem.The effectiveness of the control framework is verified by simulation.Secondly,a path planning controller based on obstacle potential field and model predictive control is designed to avoid obstacles when tracking the pre-defined path.The repulsion potential function in the traditional artificial potential field(APF)method is improved.Then,the lateral and longitudinal distances between vehicles and obstacles are normalized.What's more,obstacles are classified,and the corresponding repulsion potential field functions are designed for each kind of obstacles.To guide vehicles to avoid obstacles,the total repulsion potential field is added to the MPC objective function.The effectiveness of the algorithm is verified by simulation.Finally,to improve the tracking accuracy and stability,a lateral and longitudinal integrated control framework is proposed.According to the fuzzy control theory,the longitudinal speed is decided and controlled.The vehicle stability degree is calculated based on the sideslip angle phase plane stability theory.Then,the path tracking controller jointly controls the front wheel angle and the direct yaw moment,and adjusts the weight of tracking accuracy and stability in the control target according to the vehicle stability degree.To realize the direct yaw moment,a torque distribution controller is designed to distribute the total torque to four wheels.At last,the above parts are integrated in the lateral and longitudinal integrated control framework.The effectiveness of the control framework is verified by simulation.