Research On The Strategy And Control Method Of Intelligent Vehicle Lane Change Based On Driving Safety Field And Model Predictive Control

Vehicle driving behavior decision-making,behavior planning and behavior control,as the three core issues of intelligent vehicle control,have been widely concerned by scholars in various countries.In the process of vehicle control,suitable driving behavior decision-making not only eliminates the redundant re-planning process,but also provides a more accurate operating basis for the control layer.In order to improve the safety of intelligent vehicle lane changing,this paper proposes a lane change strategy and control algorithm based on safety field and model predictive control(MPC),simulates and analyzes the algorithm as well.First of all,in view of the possible uncertain safety problems of the global lane change path,this paper analyzes the dynamic environment change characteristics in the process of vehicle lane change,puts forward a discrete point safety path decision strategy based on the safety field,determines the minimum risk path point set in the lane-changing scene,and realizes the decision-making of safe driving behavior of intelligent vehicle lane change.Secondly,a path-speed decomposition discrete point fitting track planning algorithm is designed,and the discrete path points obtained by the decision-making layer are fitted with a fourth-order polynomial,considering the acceleration and time of lane change,with the goal of improving the comfort and the overall road traffic utilization rate,an optimization objective function is established,to achieve safe,comfortable and fast intelligent vehicle track planning.Then,using the dynamic model and the model predictive control algorithm,the control amount and the driving boundary conditions are used as constraints to establish the objective function that minimizes the tracking error,and the optimal wheel steering control input is obtained through optimal solution to achieve a safe and comfortable lane change trajectory tracking control.Finally,a joint simulation platform of Car Sim and MATLAB/Simulink is built.Using three different driving scenarios at different speeds the paper’s intelligent vehicle lane-changing decision-making,planning and tracking control algorithms were verified and compared with other security risk detection algorithms.The simulation results show that under different driving scenarios,the discrete point lane change decision-making method based on the safety field can provide a safe and real-time lane-changing strategy for the intelligent vehicle.Compared with other safety risk decision-making methods,this method has the characteristics of more comprehensive considerations and more reasonable decision-making results,which greatly simplifies and improves the safety behavior decision-making of intelligent vehicle,and provides better technical support for lane-change behavior decision-making.The path-speed decomposition of discrete point fitting track planning algorithm sits quickly to plan a safe lane-changing path,and a comfortable and efficient lane-changing speed for the intelligent vehicle.The tracking algorithm of the model prediction control realizes the track tracking control of the intelligent vehicle lane change,and can control the tracking error between the actual driving track and the reference track of the vehicle to ?0.5m.