Research On Unmanned Ground Vehicle Inspection Path Planning

With the development of the information,the functions of unmanned ground vehicle(UGV)are increasingly wealthy,playing an important role in logistics and transportation,indoor cleaning and security inspection,and other fields.Especially in the area of security inspection,UGV can replace manual review,and effectively save human resources,improveing the safety of inspection.UGV inspection path planning is one of the key research directions in security inspection.There are two problems in UGV inspection path planning: One,the coverage of inspection points is too low;The other,the efficiency of path planning between inspection points is poor in a complex environments.This thesis will focus on the above two problems.Firstly,aiming at the problem of low coverage of inspection points in UGV inspection path planning,this thesis proposes a particle swarm potential field(PSPF)algorithm for inspection point planning.When particle swarm optimization(PSO)algorithm is used to realize region coverage,it is easy to fall into the local extrema problem,which leads to the formation of large-scale overlapping coverage areas and a large number of uncovered spaces.However,due to the limitation of the PSO,particles trapped in the local extrema will be slowly,so that the particles can not escape from the local extreme value,resulting in a low coverage of the final planned inspection point.To solve the problem of low coverage of inspection points,the PSPF algorithm combines the PSO and artificial potential field method closely.It introduces the resultant calculation of the synthetic possible field method into the velocity updating formula of the PSO algorithm.When the particle falls into the local extrema,the overlapping coverage area generates repulsion,while the uncovered area generates gravity.The combined force of gravity and repulsion pulls the particle out of the current position and the local extrema.Compared with the simulation experiment of PSO,the coverage of PSPF is 25% higher than that of PSO,which proves the effectiveness of PSPF.Secondly,aiming at the efficiency of path planning between UGV inspection points,this thesis proposes a path planning method based on the ant colony algorithm of a regular hexagon grid map marked by the double coordinate method(DHAG).This method uses a regular hexagonal grid method to model the environment,instead of the square grid,to solve the problem that the square grid method will bump into obstacles when moving diagonally.The double coordinate method is proposed to mark a regular hexagon grid maps,which can reduce the complexity of data storage and improve the efficiency of the ant colony algorithm in path planning on this map.Compared with the simulation results of path planning using the ant colony algorithm on square raster maps and regular hexagonal raster maps,the DHAG method reduces the running time by 60% and 20%,respectively,proving the effectiveness of the DHAG method.Finally,this thesis realizes the map import,checkpoint planning,path planning,and patrol functions based on the PSPF algorithm and DHAG method proposed above.The modular unmanned system control platform is built according to the joint architecture for unmanned systems(JAUS),and the realized functions are deployed on the platform as the UGV inspection application module.