Research On Mapping And Localization Algorithm Of Mobile Robot Based On Reflector In Complex And Changeable Environment

At present,mobile robots have been widely used in industrial fields,which has great significance for factory automation construction.High-precision mapping and localization algorithm is the key to the efficient and intelligent work of industrial mobile robots.Limited by the complex and changeable characteristics of the factory environment,industrial mobile robots still have problems such as poor stability and low accuracy of mapping and localization,which seriously limit the large-scale popularization of industrial mobile robots.Therefore,this paper takes mobile robots in complex and changeable environments as the research object and studies the high-precision mapping and localization algorithm of mobile robots based on the light intensity characteristics of reflectors,which is of great practical significance and application value for accelerating the application of industrial mobile robots.The main research contents of this paper are as follows:According to the complex and changeable characteristics of the industrial mobile robot operating environment,the functional requirements of map building and localization algorithm are analyzed.According to the functional requirement analysis,the design and construction of the mobile robot experiment platform,including mechanical structure,software system,sensor units,etc,is complete.The basic modules of the SLAM system,including the coordinate system,mobile robot motion model,lidar observation model,etc.,are built to provide the basis for subsequent experiments.Aiming at the shortcomings of the ICNN algorithm with low data association success rate and the JCBB algorithm with high time complexity,an improved data association algorithm is proposed to improve the accuracy of the secondary data association by dividing the data association into two sessions and using the results of the initial data association to correct the inter-submap poses.Compared with JCBB algorithm,the execution time of this algorithm is reduced by 18.0ms on average,and its success rate is basically maintained around 95%,which is 11.0%higher than that of ICNN algorithm.The results show that this algorithm can effectively reduce the time complexity of data association algorithm while maintaining a high success rate of data association.To address the problems of low accuracy and high time complexity of existing loop closure detection algorithms,a fast loop closure detection method based on octree is proposed.It has a high accuracy of 99.28%at a recall rate of 63.65%,and the execution time of a single loop closure detection algorithm is maintained below 10 ms.For the problems of high time complexity and large cumulative error in the process of map building,the SLAM algorithm,using a combination of submap strategy and graph optimization method,is proposed.The experimental results show that the average error in xy direction of the reflector is 6.8mm and 9.6mm,which is 13.2mm and 19.2mm less than that of the I-SLSJF algorithm,proving that this algorithm can effectively improve the accuracy of reflector map building in large-scale environment.In order to solve the problem that the global localization takes a long time in the largescale reflector map and the matching error may occur,a fast global localization algorithm based on octree is proposed.The average time of global localization in the 2500m2 environment is 73.9ms,which is 67%less than the average time of mature products in the market.The xy direction error of the global localization results is less than ± 10mm,and the heading angle error is less than ±0.5°.It shows that the global localization algorithm can effectively improve the speed and accuracy of the global localization of the reflector on the premise of ensuring the localization accuracy to meet the industrial localization requirements.An position tracking algorithm for reflector based on EKF is proposed,which can meet the localization accuracy requirements of the industrial mobile robot with a repeat localization accuracy of less than 10 mm in xy direction and less than 0.5° in heading angle error and has better stability in the position tracking process.In order to solve the robot abduction problem in the reflector map,an evaluation criterion for localization failure is proposed which can accurately identify the mobile robot abduction situation.By combining this with the proposed fast global localization algorithm based on octree,it can effectively solve the robot abduction problem and improve the stability of the localization algorithm.