Research On Emergency Active Collision Avoidance Control Algorithm For Autonomous Vehicles

With the increasing number of vehicles year by year,road traffic problems are becoming more and more serious,people pay more attention to road traffic safety,and more and more active safety technologies have been promoted in the market.At present,most researches on vehicle longitudinal collision avoidance systems use collision warning or emergency braking to avoid collisions.However,when the vehicle is in special circumstances such as high speed and slippery roads,it is more effective to use emergency steering to avoid collisions.In order to further improve vehicle safety,this paper focuses on the research on vehicle longitudinal and lateral active collision avoidance control methods,analyzes and designs collision avoidance decision-making,collision avoidance path planning.And based on the automatic driving simulation software,the designed control strategy is tested and verified.The specific research content of the paper is as follows:（1）This paper summarizes and standardizes typical dangerous traffic scenarios by investigating and analyzing the research literature,traffic accident reports,and traffic regulations related to vehicle collisions.Taking the established vehicle safety distance model as the main evaluation basis for vehicle collision risk,combined with the driver’s takeover strategy and the method of judging the collision risk of oncoming vehicles,an emergency active collision avoidance decision-making algorithm is constructed.（2）Aiming at the emergency steering collision avoidance operation in the decision algorithm,this paper puts forward five specific requirements for emergency steering collision avoidance path planning,By introducing the path planning algorithm based on quintic polynomial,this paper leads to the multi constraint collision avoidance path planning optimization method considering the vehicle traveling area.In the process of planning,combined with the position prediction of moving obstacles and the analysis of collision danger time,an optimization function considering kinematic constraints and driveable area constraints is established.The collision avoidance path obtained by optimizing the solution can meet the requirements of collision avoidance and take into account the riding comfort of the driver and passengers.（3）In order to meet the requirements of upper-level decision-making planning,this paper studies the vehicle lateral and longitudinal motion control algorithm.In the aspect of longitudinal control,the feedforward control quantity is established by calculating the driving resistance and the inverse model of the braking and driving system,and combined with the feedback control quantity based on the PID control algorithm,the acceleration tracking control algorithm of vehicle feedforward and feedback is established.In the aspect of lateral control,a prediction model of vehicle-road pose deviation is established based on the simplified vehicle model,and the MPC path tracking controller is designed on this basis.（4）In this paper,a co-simulation environment is built by Matlab/Car Sim/Prescan software,and the integrated simulation verification of the proposed decision-making algorithm,planning algorithm and motion control algorithm is carried out.Considering the problems of sensors in application scenarios,this paper further processes the information obtained by the sensor.According to the summarized typical dangerous traffic scenarios,the corresponding test conditions are set to verify the collision avoidance effect of the proposed control strategy in different dangerous scenarios.