Research On Obstacle Avoidance Strategy Of Intelligent Vehicles In Vehicle Network Environment

In recent years,the rapid development of the national economy and the increase in population have led to an almost exponential growth in car ownership,and urban traffic problems such as traffic jams and traffic accidents need to be resolved.Therefore,it is of great theoretical significance and practical application value to carry out relevant technical research on assisted driving technology to improve road traffic safety.However,current research on the method of automatically changing lanes is very limited,relying only on its own sensors to obtain external information.During the entire lane change process,it is assumed that the surrounding vehicle state is static and contradicts the complex and variable actual road conditions.Therefore,this paper proposes an automatic lane change strategy based on vehicle network information.Smart cars use the Internet of Vehicles and sensors to sense the driving environment.The sensors include millimeter-wave radar,GPS,camera and laser radar.These raw data are combined by the “sensor data fusion module”,and the car network transceiver unit receives the real-time information of the car network.Get dynamic and static obstacle information.Based on the real-time information of obstacles,the system plans a safe,comfortable and efficient driving track,and then tracks and controls this track.This paper first analyzes and summarizes the local path planning method and path tracking control method at home and abroad,and selects the improved polynomial theory as the method of obstacle avoidance path planning.In the path planning based on polynomial theory,the surrounding road information and obstacle information The motion state is analyzed,and the minimum safety distance is selected as the performance parameter of the improved polynomial path to ensure the traffic efficiency and safety.After the optimal lane change trajectory is obtained,the MPC-based path controller is used to track the lane change trajectory..The main research contents of this paper are:(1)Formulation of the lane change strategy.The decision-making strategy and route direction of intelligent vehicle lane change are studied.Based on the actual lane change behavior characteristics,lane change judgment analysis mechanism and minimum distance model,the main lane change decision algorithm is designed.(2)The estimate of the TLC of the crossover time.In this paper,the geometric function method is used to establish the TLC estimation model for the cross-line situation in different scenarios.In order to ensure the validity of the model,the relevant geometric parameters of the vehicle body are considered.The MATLAB simulation software is used to simulate and analyze the parameters affecting TLC,including yaw angle,lateral acceleration,road bending radius and lateral displacement.The simulation proves that the model can predict and estimate the value of TLC in real time.(3)Determination of the optimal lane change trajectory under the condition of network association.The polynomial method is used to carry out the path planning,and the objective function is established on the basis of this,and the optimal lane change is obtained according to the constraint conditions.Finally,the minimum safety distance model is added to the lane change condition to ensure the safety.The required parameters are real-time car network data provided to effectively avoid the danger caused by sudden situations(such as sudden intrusion,changes in surrounding speed and sudden lane change,etc.),thus ensuring the safety during the entire lane change process.(4)Path tracking based on MPC controller.The MPC-based control system is used to control the front wheel steering and single wheel torque to track the required collision-free reference trajectory and to verify the relevant parameters of the lane change trajectory by simulating the pilot.