Survey Of Path Planning Algorithms For Autonomous Vehicles

Autonomous vehicles play an important role in promoting future economic development,environmental friendliness,and convenience in life.They have been valued and developed rapidly by countries around the world.Path planning is an important part of autonomous driving technology and a prerequisite for safe driving.At present,the path planning algorithms for autonomous vehicles face many difficulties,such as poor ability to cope with complex environments and poor robustness in handling dynamic obstacles,which seriously affect the driving safety and planning efficiency of autonomous vehicles.Therefore,the path planning for autonomous vehicles has important theoretical significance and research value.Based on the characteristics of autonomous vehicles,this paper analyzes and summarizes the current commonly used path planning algorithms.Combining with the vehicle kinematics model,it selects the appropriate algorithms for global path planning and local path planning algorithms,and improves and optimizes the algorithms according to their shortcomings.The simulation test is carried out from the aspects of planning path length and quality,algorithm calculation efficiency and obstacle avoidance ability,etc.The experiment proves that the improved algorithm proposed in this paper is feasible and effective.The main work of this paper is as follows:(1)The kinematics analysis and modeling of the vehicle were carried out,and the two-dimensional occupation grid environment model was established,which laid a foundation for the following research work of path planning algorithm.Secondly,the ant colony algorithm is used as a global path planning algorithm,and an improved ant colony optimization algorithm is proposed.In order to improve the planning efficiency of traditional ant colony algorithm,the quality of path planning and the feasibility of paths,the ant colony algorithm's pheromone concentration,state transition probability,and the introduction of heuristic functions were improved and optimized.The comparison experiment between the improved ant colony optimization algorithm and the traditional ant colony algorithm shows that the planning speed of the improved ant colony optimization algorithm is greatly improved,the corners of the path are reduced,and the path length is decreased.(2)The ant colony algorithm(ACO)is taken as the global path planning algorithm and an improved ant colony optimization algorithm is proposed.Aiming at improving the planning efficiency,path planning quality and path feasibility of the traditional ant colony algorithm,the algorithm was improved and optimized from the directions of pheromone concentration,state transition probability and introducing heuristic function.Through the comparison experiment between the improved ant colony optimization algorithm and the traditional ant colony optimization algorithm,the planning speed of the improved ant colony optimization algorithm is greatly improved,the corner of the path is reduced,and the path length is reduced.Finally,a simulation environment for the improved doublelayer path planning algorithm is set up in the Matlab Automated Driving System Toolbox,and vehicle movement is simulated in the road environment to verify the feasibility and effectiveness of the improved double-layer path planning algorithm.(3)Dynamic window algorithm is used in local path planning.Aiming at the problem that the dynamic window algorithm has low planning ability for large and complex obstacles and is easy to get into deadlock,an improved two-layer path planning algorithm is proposed.The global optimal path planned by the ant colony optimization algorithm is taken as the path guide of the dynamic window algorithm,the scroll pane is added,and the direction Angle adaptive adjustment method is specified.The improved two-layer path planning algorithm performs well in complex environments and the planning time is obviously shortened.(4)Matlab platform and the Automated Driving System Toolbox were used to build a simulation environment for the improved two-layer path planning algorithm.Simulation verification of the algorithm was conducted by simulating vehicle driving in the road environment to prove the feasibility and effectiveness of the improved two-layer path planning algorithm.Finally,the algorithm is verified by an autonomous experimental vehicle.