Research And Application Of Dynamic Path Planning Algorithm Based On Probabilistic Road Map

Path planning is a process that generates a path from a source location to a destination location.Path planning combines its applicable objects and application scenarios to find an optimal path without collisions in the presence of obstacles according to certain criteria(such as path length,running time,etc.).In the practice of production and life,mobile robots and smart vehicles are usually in a dynamic environment that may change in real time,such as the presence of movable obstacles or tracking moving targets.This requires mobile robots or smart vehicles to have the ability of dynamic path planning.As a basic skill for mobile robots and intelligent vehicles to carry out various tasks,dynamic path planning has always been in the position of important research topics.The classic~*A algorithm can always find the optimal path with the lowest cost,but the algorithm is only suitable for static path planning,it takes a long time to run,and its work efficiency is low.Therefore,it is necessary to improve the traditional~*A algorithm to improve its operating efficiency.The probabilistic roadmap method uses a small number of sampling points to determine a feasible solution through random sampling.Although the solution is not optimal,its planning time is greatly shortened and the path is close to optimal.Therefore,in order to improve the planning efficiency of theA~*algorithm,a hybrid path planning algorithm can be developed by combining theA~*algorithm with the optimal path and the probabilistic roadmap method with the fast search capability.The distributed consensus algorithm uses control theory tools to provide new ideas for path planning.But this algorithm is only suitable for grid maps with large calculation and long running time,and the results of the work are limited to a static environment,and it may fail in a dynamic environment.This paper mainly has three research contents:First of all,theA~*algorithm and the probabilistic roadmap method are combined to perform static path planning based on a randomly generated work map to improve the~*A algorithm and improve its work efficiency,greatly shorten the time spent on operation and compare the performance with the distributed minimum consensus algorithm based on the probability roadmap.In addition,the distributed minimum consensus algorithm based on probabilistic roadmap is used to solve the static path planning problem,which significantly improves the efficiency of static path planning.Secondly,in order to solve the problem that the minimum consensus algorithm with a large amount of calculation and long running time in the grid,this paper proposes a hybrid algorithm that combines the probabilistic roadmap method and the minimum consensus algorithm with a grid,and changes the grid map to a random map to improve its operational efficiency,and it is used to solve maze-type problems,avoid U-shaped obstacles and dynamic planning of mobile robots,such as tracking moving targets and tracking moving targets with changing obstacles.Finally,this paper improves the biased minimum consensus algorithm based on the specific problems that smart vehicles may encounter during operation,and then combines the improved algorithm combines with the probabilistic roadmap method to form a hybrid dynamic path planning algorithm for smart vehicles.Considering the road conditions(congestion index,whether an accident occurred,etc.)and the Euclidean distance,the biased minimum consensus algorithm was improved to meet the needs of smart vehicles to plan the path in real time,so that smart vehicles may be in a dynamic environment that may change in real time can always keep driving on the best path.For example,during the driving process,the road ahead is congested due to traffic accidents.Intelligent vehicles can also re-plan the path in time to avoid congestion by changing the forward route.