Research On Motion Planning Of Unmanned Vehicles In Freeway Environment

In order to reduce the traffic accidents caused by the man-car-road link imbalance caused by human factors,unmanned vehicle have received wide attention worldwide.As one of the common key technologies of unmanned vehicle systems,motion planning has become a hot topic in this field.Most existing motion planning algorithms only consider constraints such as travel time and moving distance,and do not involve road boundary and vehicle dynamics constraints.When the environment is constructed,the vehicle state prediction model used is relatively single,and the state of the vehicle cannot be accurately estimated.For this reason,this paper takes the unmanned vehicle traveling under the highway scene as the research object,proposes a motion planning algorithm based on the improved visibility algorithm,and based on this,adds the random reachability prediction model to estimate the driving status of moving vehicle to improve the accuracy of the algorithm.The specific research content is as follows:(1)Combine the function,principle and structure of autonomous decision-making system for autonomous vehicles,analyze the organizational structure of autonomous decision-making systems,and summarize the design criteria of behavioral decisionmaking subsystems by analyzing the characteristics of human driving behavior;The driving behavior decision making subsystem is modeled and the local target point is selected according to the driving intention determined by the hierarchical state machine.(2)In order to estimate the motion of the vehicle,a Markov chain is used to obtain the vehicle’s random reachability model.On this basis,the main factors affecting the random accessibility of vehicles are analyzed to improve the accuracy of the vehicle’s random reachability model.(3)Analyze the design criteria and composition of the motion planning subsystem;establish a new sampling algorithm and add B-spline optimization function and path evaluation function to improve the traditional viewability algorithm;The reachable set model was added to improve the viewability algorithm to obtain the final motion planning algorithm.(4)Carry out motion planning in static environment,dynamic environment and mixed environment,use Carsim simulation software to verify the planning results;compare the planning results with the planned results of RRT algorithm.The results show that the planning path obtained in this paper has better comfort and safety.