Research On Intelligent Vehicle Path Planning

Path planning is an important part of the field of intelligent driving research.It is also the prerequisite for intelligent vehicles to complete driving tasks.According to the requirement of the project,this paper studies an improved rapidly-exploring random tree(RRT)algorithm method with the fourth-order Bezier curve in path planning.First,this paper studied a route planning method based on improved RRT algorithm.After the analysis of the advantages and disadvantages of basic algorithm principles,the direction of improvement can be clearly identified.Therefore,the basic RRT algorithm has been improved by introducing three aspects such as the angle constraint strategy,the goal bias strategy,and the neighborhood reconstruction strategy.This step obtained a reference route that satisfies the basic obstacle avoidance constraints.Second,this paper studied a path optimization method based on a fourth-order Bezier curve.This method can divide the reference path with five control points,due to these points several fourth-order Bezier curves that satisfied sufficient constraint,target position constraint,continuous constraint and smoothness constraint are determined.From these candidate curves,an optimal route can be selected according to the obstacle quantitative reference evaluation.Third,this paper studied a speed control method.By analyzing the vehicle kinematics model,if vehicle does not slide away,vehicle speed and acceleration are continuously bounded,this method can control steering wheel execution and front wheel steering angle to plan reasonable speed.With the support of company’s real vehicle test program,a real vehicle test was performed on a fully enclosed intelligent driving test road.According to the simulated data in the computer and the real vehicle verification data,the validity and feasibility of the study can be proved in this paper.