Research On Path Planning Method Of Patrol Robot

Patrol robot is an automatic,semi-automatic or human controlled robot to help complete the security work.One of its main technologies is path planning.In a complete planning task,the robot must face both the outdoor environment and the indoor environment,the difference between the two environments is very big.Therefore,this research divides the task of patrol robot path planning into outdoor environment path planning and indoor environment path planning.In the outdoor environment,the patrol robot needs to face a wide range of global map,so the main problem of planning task is to control the path length and time loss of the robot.In order to solve these problems,this paper proposes a method to optimize the map by using traffic flow prediction model while helping the robot to build outdoor map based on GPS technology.Then,the simulation experiments of probabilistic search algorithm,heuristic search algorithm and stochastic search algorithm in outdoor map are compared and analyzed,and the optimal probabilistic search algorithm is selected to complete the outdoor path planning task.The indoor environment is narrow and sudden,so the main problems faced by robots are the rapid construction of indoor map and the avoidance of dynamic obstacles.For the construction of map,this paper uses the implicit definition method based on state space to build local environment map.In order to avoid obstacles,robots need to be able to realize dynamic path planning,which requires high real-time performance of planning algorithm.Therefore,this paper studies the random search tree algorithm with strong real-time performance,and proposes a re planning strategy based on "angle difference detection between robots and obstacles" to achieve obstacle avoidance.In addition,the multi robot path planning task is studied to solve the problem that the robot needs to execute multi tasks in parallel.At this time,many robots work in the same environment,so the random search tree algorithm is optimized by using the method ofmulti robots sharing path nodes.