Research On Construction Method Of Topology Map Of Nuclear Power Station Environmental Based On Establishment Of Landmarks

With the shortage of traditional resources such as petroleum and coal and the urgent need for clean energy development in various countries,nuclear energy as a renewable clean energy has been listed as the strategic focus of energy development in the country,and nuclear power robots serving the nuclear power environment have emerged as the times require.This project is based on the project of National Natural Science Foundation of China "Research on Robot Adaptation and Efficient Operating Method for Nuclear Power RCV"(61473113).Accurate mapping of working environment is the basis of autonomous operation of nuclear power robots in nuclear power environment.In this paper,a large-scale and complex space topological map construction method and landmark feature extraction of point cloud data are studied.The main contents and innovative achievements of this paper are as follows:Firstly,to eliminate the influence of camera distortion on image quality,the camera is calibrated with Open CV vision library and PCL point cloud library,and the point cloud data are collected in the simulated nuclear power environment.In order to improve the processing efficiency of point cloud data,an improved bilateral filtering algorithm is proposed to denoise the point cloud.Then,the best connected region method is used to partition the point cloud data into cubes and filter and remove outliers.Secondly,by analyzing the problems existing in current topographic maps,a method of constructing topographic maps based on landmarks is proposed.In this paper,the creation of landmarks is studied.Aiming at the identification of obstacles and topological nodes,a method of feature extraction of cylindrical landmarks based on RANSAC(Random Sample Consensus)is proposed.The position of landmarks is represented by cylindrical axis and coordinates,and the radius of the cylindrical is expressed by the size of the landmarks.The extracted landmarks are combined with nodes to provide obstacles for navigation of mobile robots.Obstacle identification information.Thirdly,aiming at the construction of the topological map of the target space,this paper proposes a special state-based mobile strategy for nuclear power robots,which achieves the creation and selection of the topological nodes.Through the closed-loop detection of the topological map,the accuracy of map construction is improved,and the map nodes can be updated by observing the belief state.In this paper,the method of multi-local map fusion is used to further improve the accuracy and robustness of map construction.Based on the ICP(Iterative Closest Point)algorithm,each local map is fused.When the error between two iterations is less than the set threshold,the iteration is completed.Finally,an experimental platform for simulating nuclear power environment is built to verify the proposed algorithm.The experimental results show that the proposed filtering algorithm can effectively reduce the noise in point cloud data.Compared with traditional methods,the proposed landmark extraction method based on RANSAC improves the operational efficiency and accuracy.The number of times and success rate of building topological map show that the method of building topological map in this paper is feasible and effective.