Design And Implementation Of Key Technology For Inspection Robot

With the rapid development of robotics,robots have gradually replaced humans to perform highly repetitive and complex and dangerous labor.Especially when performing inspection tasks,manpower is often required to repeat operations and there is a possibility of danger,and Patrol Robot(PR)has higher inspection efficiency while saving manpower and material resources.Most of the existing inspection robots are fixed in various areas of the inspection scene,and only rely on image technology to complete patrol monitoring.It requires a lot of deployment in mission scenarios,has poor autonomy,and is limited by terrain.Aiming at the problems of poor autonomy and inability to adapt to complex terrain of existing inspection robots,this paper studies the key technologies of Simultaneous Localization and Mapping(SLAM)and autonomous path planning in the inspection process,proposes an improved SLAM technology,integrates the improved global and local path planning algorithms,and designs an inspection robot system that can conduct autonomous inspection,so as to achieve the purpose of intelligent inspection without being limited by terrain.The results of this paper include the following:Firstly,aiming at the problem that the existing particle filter(Rao-Blackwellised,RBPF)algorithm has insufficient positioning accuracy and relatively single particle form,this paper proposes an improved SLAM algorithm,which integrates the observation vector at the current moment in the process of sequential importance sampling(SIS),and then samples the improved recommended distribution function.In this way,the prediction of the robot’s real-time pose state and the accuracy value of map construction are significantly enhanced.Through comparative analysis,the results show that the new improved algorithm proposed in this paper has more advantages in the accuracy of lidar registration accuracy and state evaluation.Secondly,in view of the slow retrieval speed and tortuous calculation of the A*algorithm,this paper proposes an improved A* algorithm,which significantly improves the global retrieval speed and greatly reduces the time of node reciprocating search by increasing the weight coefficient of its heuristic function and integrating the node distance into the exponential decay function.The A* algorithm designed and updated in this paper is sampled and matched with the Dynamic Windows Approach(DWA)local programming algorithm,and the global optimal path is obtained by the adjusted evaluation function.By analyzing and comparing with the traditional A* algorithm,the search time of the calculated path of the improved A* algorithm in this paper is shortened by half,and the number of computing nodes and the number of related inflection points are greatly reduced.Finally,the inspection robot inspection system based on Turtlebot is built for the two different working conditions of outdoor and indoor,and the improved SLAM mapping method and the fusion path planning algorithm are compared and tested,and the results show that the inspection robot inspection system meets the requirements of high-precision mapping and autonomous path planning.