Research On The Path Planning And The Tracking Control Algorithm Of The Intelligent Vehicle In Complex Environment

At present,countries around the world have paid great attentions to road traffic problems,It has an important impact on the personal safety of drivers and pedestrians and the national economy.It is foreseeable that in the future,the integrated intelligent transportation system will be widely applied,and the intelligent vehicle is a key component of the system.With the continuous development of intelligent vehicles,some relative intelligent driving assistant technologies have became mature,such as the automatic emergency braking technology and the adaptive cruise technology,which can assist the driver in the general driving environment to reduce the burden.The higher level of intelligent driving requires intelligent vehicles to automatically avoid obstacles in complex environments with the most optimal strategy.Therefore,it is necessary to further study the path planning method and path trackingcontrol strategy of intelligent vehicles in the complex traffic environment in order to achieve a higher level of intelligent driving.This paper is supported by National Key Research and Development Program Key Special Projects for New Energy Vehicles "Key Technology Research and Demonstration Operation of Electric Automatic Driving Vehicle"(Number: 2017YFB0102600).This paper studies the path planning and path trackingcontrol of intelligent vehicles in complex environments,including vehicle kinematics and dynamics modeling;dynamic path planning of local obstacle avoidance for intelligent vehicles;path tracking control strategy based on Model Predictive Control(MPC)and co-simulation verification of the proposed research.The specific contents of the paper are as follows:(1)Firstly,this paper studies path planning method for intelligent vehicles to avoid obstacles in complex environment(this paper refers to the road environments with more static and dynamic obstacles,typical working conditions such as parking lot,downtown area,etc.).Aiming at the problems of traditional artificial potential field(APF)algorithm,such as target unreachability,local minimum and path jitter,an improved artificial potential field algorithm is proposed.On the basis of traditional artificial potential field method,a comprehensive constraint model of artificial potential field method is established by adding road boundary constraints,dynamic and static obstacle constraints and vehicle motion characteristics constraints.Then,the model is validated by simulation experiments in different complex environments(only static obstacles,dynamic obstacles far away,dynamic obstacles near),and the planned path is also validated by curvature analysis.The experimental results show that the proposed algorithm can plan the obstacle avoidance path under the condition of both local dynamic and static obstacles,and the curvature of the planned path is continued and smooth.(2)To ensure track the planned path accurately,a MPC-based vehicle path trackingcontroller based on the hierarchical control theory is designed in this paper,including motion control layer and execution control layer.Based on the three-degree-of-freedom dynamics model including longitudinal,lateral and yaw of the vehicle,a MPC motion controller is established when using the lateral velocity,the longitudinal velocity,the yaw angle,the yaw rate and vehicle position as state variables.The output variables are longitudinal force and lateral force of front and rear tires in body coordinate system;In the executive control layer,the longitudinal and lateral forces of front and rear tires from the upper layer input in the body coordinate system are converted into those in the tire coordinate system.and then the rotation angle and the driving torque of each wheel are obtained based on the inverse tire force model.(3)In order to verify the effectiveness of the proposed path planning and tracking controll algorithm,a co-simulation platform is built by Matlab/Simulink and Car Sim software.Experiments were carried out at different vehicle speeds and different adhesion coefficients.The experimental results show that the proposed path planning and the tracking control algorithm has good simulation effect,and has strong obstacle avoidance,adaptability and control performance.