Research On Vehicle Lane Changing And Obstacle Avoidance System Based On Model Predictive Control Algorithm

With the healthy and rapid development of the national economy,the number of cars is also increasing,and the problem of driving safety is more concerned.The active obstacle avoidance system of vehicles has always been the focus of active safety technology research.When there is an obstacle in front of the vehicle that may cause a collision accident,the driver can take two different operation methods of braking or steering to avoid the accident.However,for poor road adhesion conditions and a relatively high speed relative to the obstacle,Braking may not completely avoid collisions and will reduce road utilization.In order to solve the above problems and achieve the purpose of accurate obstacle avoidance,this paper designs a vehicle active lane avoidance obstacle avoidance system from the perspective of steering obstacle avoidance,and makes an in-depth study on the decision strategy and lateral motion control of the system design.First of all,system modeling is carried out according to the system’s functional requirements and research needs.A three-degree-of-freedom vehicle dynamics model is established according to the Newtonian mechanics analysis method.The magic tire model is used to simplify the nonlinear vehicle dynamics model as the basis for model prediction.Then,a lane change and obstacle avoidance model is established for the system function,and the longitudinal braking safety distance and lane change safety distance are comprehensively considered,and a safe distance model is established to judge the driving state of the vehicle,which is used as an indicator for the lane change warning decision strategy.Secondly,design the overall architecture of the active lane change obstacle avoidance system,including system functions and system solutions,and focus on the research of decision strategy and trajectory planning of the control system.The lane change warning decision is based on obstacles and environmental information monitored in real time by the perception layer,combined with the safety distance model and driver status recognition to determine the vehicle’s driving safety and lane change feasibility.The trajectory planning part chooses quintic polynomial trajectory planning by comparing with common methods.This method has a small amount of model calculation,and the shape of the planned trajectory is closer to the actual trajectory of vehicle steering and obstacle avoidance.Then the trajectory planning parameters are determined according to the lateral stability constraints and road boundary constraints.Then,aiming at the problem of obstacle avoidance trajectory tracking,a design method of trajectory tracking controller based on model predictive control(MPC)is proposed.The nonlinear dynamic model after linear discretization is used as the predictive model of the controller,combined with the constraints of the vehicle lane change process The conditions are optimized for online rolling,and the front wheel angle of the vehicle,that is,the control amount of the steering control,is solved to control the vehicle to track the planned trajectory of obstacle avoidance.Using d SPACE ASM and Matlab/Simulink to complete the joint simulation of different speeds,different road adhesion conditions and obstacles in different states,etc.,to verify the correctness of the algorithm of the lane change obstacle avoidance system.Finally,based on d SPACE hardware and software and real sensors,the hardware-in-theloop test platform was used to test the vehicle lane changing obstacle avoidance at different speeds.The test results verified the real-time and effectiveness of the lane change obstacle avoiding controller.