Research On Path Planning Of Reconnaissance Unmanned Vehicle In Extremely Dangerous Environment

Reconnaissance unmanned vehicle play an important role in extremely dangerous environments such as fire smoke tunnels,leaking nuclear power plants,chemical plants full of toxic and harmful gases,and fire scenes.In order to enable reconnaissance robots to complete the established tasks accurately,it is necessary to have path planning technology as the basis.Path planning technology is the key to navigation of mobile robots,so the research on path planning technology is of great significance for intelligent mobile robots.However,most of the current research work on path planning technology is still at the theoretical stage,and various algorithms have different advantages and disadvantages.For example,A * algorithm may plan longer paths and more turning points;The particle swarm optimization algorithm may fall into the local optimal problem in the later period.This paper studies the defects of particle swarm optimization algorithm.The main research work is as follows:(1)The actual working space of the reconnaissance unmanned vehicle is modeled.The modeling of the working environment of the reconnaissance unmanned vehicle mainly includes two-dimensional grid map modeling and three-dimensional space simulation modeling.The workspace is a static space,which only contains static obstacles.(2)Analyze the shortcomings and reasons of particle swarm optimization algorithm in path planning,and on this basis,study how to effectively solve the problem of local optimization and the contradiction between convergence speed and diversity,finally make the algorithm have faster convergence speed and higher solution accuracy,and effectively balance the contradiction between exploration and utilization.Analyze and test the selection of each parameter in the basic particle swarm optimization algorithm,summarize the impact of each parameter on the algorithm performance and determine the reasonable setting range of each corresponding parameter,so as to pave the way for parameter optimization and improvement.In order to solve the problem that PSO may fall into the local optimal solution,the Levy flight method is introduced to improve PSO,and the iterative formula of PSO is modified.This can avoid the local convergence of particle swarm after a certain number of iterations,so as to improve the convergence speed and find the global optimal solution.(3)The improved path planning algorithm is tested in 2D Matlab and 3D environment respectively.The 3D environment test includes virtual 3D environment and real environment application test.First,build the hardware platform of the reconnaissance unmanned vehicle and set a series of parameters.Then,the reconnaissance unmanned vehicle uses the improved algorithm to search for an optimal path without collision,avoiding the obstacles set and successfully reaching the target point.Experimental results show the effectiveness of the algorithm.