Collision Avoidance Trajectory Planning Of Intelligent Vehicle Based On The Trajectory Prediction Of The Front Vehicle In The High-speed Scene

At present,with the rapid development of Expressway in China,traffic accidents are also increasing.Compared with the complex traffic conditions of urban roads,the expressway is more regular and the vehicle condition is simpler.Therefore,in recent years,the solution of automatic driving in the high-speed scene has become a research hotspot of universities and enterprises.It is a common traffic behavior to change lanes and avoid obstacles in high-speed scene.Intelligent vehicles have a high speed in the process of changing lanes and are prone to emergency situations.The track prediction of surrounding traffic vehicles,the planning of lane changing speed of main vehicles,and the safe distance from surrounding traffic vehicles are all of great significance for the smooth lane changing and obstacle avoidance of intelligent vehicles in high-speed scene.Based on the project of "Research on multi-objective coordinated control method of vehicle emergency collision avoidance with independent authority under emergency conditions" of NSFC(Project No.51875184),this paper takes the process of lane changing and obstacle avoidance on expressway as the research object,and studies the problem of intelligent vehicle collision avoidance path planning under high-speed emergency conditions,including the problem based on static state The main vehicle collision avoidance path planning of the obstacle vehicle location,the path prediction of the obstacle vehicle,the main vehicle collision avoidance path planning based on the predicted path of the obstacle vehicle,and the simulation of the main vehicle collision avoidance path under different driving conditions of the front vehicle based on MATLAB software.The specific contents of this paper are as follows:(1)The Bezier curve is used to plan the main vehicle collision avoidance path.The vehicle speed and acceleration are obtained by time derivation.The planned path is evaluated by vehicle dynamics.The objective function including safety,stability and traffic efficiency is constructed in the process of vehicle collision avoidance.Sequential quadratic programming is used The location parameters of Bezier curve’s control points are optimized by programming(SQP)method,and the optimal collision avoidance trajectory is obtained;(2)The hidden Markov model(HMM)and Gaussian mixture model(GMM)are used to establish three driving intention models,which are left lane change,right lane change and lane keeping.Fusion of motion model and driving intention prediction to predict the future time-domain trajectory of vehicles;(3)When planning the obstacle avoidance trajectory of the main vehicle,the dynamic motion position of the front vehicle in the future time domain is considered,the uncertainty of the predicted trajectory of the traffic obstacle vehicle is analyzed,the concept of collision risk probability is proposed,and the collision risk probability evaluation index is added to the objective function to optimize the solution to obtain the vehicle collision avoidance trajectory,which is carried out in the same scene with the planning algorithm results based on the static position of the front vehicle contrast;(4)Based on the simulation of MATLAB software to verify the track planning process of the main vehicle to avoid obstacles when the front vehicle changes lane in typical straight road scene,and analyze the track planning and speed planning results of the main vehicle in the process.