Research On Intelligent Vehicle Driving Behavior Decision And Motion Planning Control

As an important part of intelligent transportation system,intelligent vehicles have important significance in reducing traffic congestion,ensuring driving safety and improving ride comfort.Although smart vehicles have developed over the years,its key technologies are still insufficient at this stage,and their performance and efficiency still have the potential to continue to improve.This paper focuses on the four aspects of driving behavior decision-making,motion planning,tracking control and real vehicle experiment of intelligent vehicles.(1)Research on intelligent vehicle driving behavior decision system based on deep reinforcement learning theory.Firstly,based on the Markov decision process,the intelligent vehicle driving behavior decision model is constructed from the aspects of vehicle driving safety,comfort and driving efficiency.Secondly,the deep Q network is used to solve this model.Finally,the simulation experiment is carried out.The results show that based on the decision-making result of the driving behavior decision-making system,the intelligent vehicle can reasonably achieve severe acceleration,gentle acceleration,uniform speed,gentle deceleration,severe deceleration and lane change.(2)Study the motion planning method of APF-RRT algorithm based on maximum corner constraint.Firstly,aiming at the motion planning problem of intelligent vehicles,based on the traditional RRT algorithm,the gravitational idea is introduced to improve the convergence speed of the algorithm.Secondly,based on the intelligent vehicle non-integrity motion model,the generation area of the new node is limited.Finally,use Cubic B-spline curve smoothes the planning curve.The simulation experiment is carried out in the complex obstacle environment.The experimental results show that the algorithm can quickly generate a safe and smooth path that satisfies the kinematic constraints of the vehicle.(3)Research on intelligent vehicle tracking control methods.Firstly,the intelligent vehicle dynamics model is established,and the state space equation of intelligent vehicle tracking control is obtained.Secondly,the tracking controller is designed based on the model predictive control algorithm and vehicle dynamics constraints.Finally,the CarSim\Simulink joint simulation platform is established.Path tracking is carried out for double-shifting conditions under different vehicle speeds and different road attachment conditions.The simulation results show that the tracking controller has good tracking effect.(4)Real vehicle experiment of path tracking control.In the campus environment,a real vehicle path tracking experiment was conducted.Experimental results show that the tracking controller has a good tracking effect.