Research On Simultaneous Localization And Mapping Algorithm For Mobile Robots Based On Gmapping

With the rapid development of science and technology,mobile robots have been widely used in the fields of manufacturing,agriculture,navigation and aerospace.Simultaneous localization and mapping(SLAM)technology is a key technology for mobile robots to successfully complete tasks in various environments,and improving the accuracy of positioning and mapping is of great significance for achieving intelligent mobile robots and wider applications.This article will conduct research based on indoor mobile robots,and the main research contents include:1.Aiming at the problems of unsmooth,thick,and uneven lines in the map created by the Gmapping algorithm,and the inconvenience of transplanting the PCL library filtering algorithm for application,a method for optimizing front-end point cloud data based on voxel filtering is proposed.Voxel filtering can dilute point cloud data without destroying the geometric structure of the point cloud itself.After testing,this scheme can filter out redundant points,achieving the effect of thinning the wall and making the lines even,smooth,and elongated.2.In view that the heavy medium frequency sample of Gmapping would worsen the rate of particles degenerate,which would affect the precision of drawing,a re sample strategy based on division management was proposed.The strategy no longer directly sample from the current particles,but divided the particles into two sections according to the weight of the two sections.After that,the total number of the particles that were sample was the final result of this round of sample.After the test,the particles with small weight were delayed to be eliminated,alleviating the rate of their degenerate,and the precision of drawing was improved.3.Aiming at the problem that the maps built by Gmapping are difficult to straighten,a map straighten algorithm based on straight line scoring system is proposed.This algorithm designs a straight line scoring system using Canny edge detection and probabilistic Hough transform.Based on the insufficient map environment information in a period of time when the mobile robot initially constructs the map,the wall straight lines can be screened out,and the map can be rotated based on this,so as to achieve the alignment of the map.Experimental results show that the algorithm can realize the map alignment in any complex environment,and has good robustness,which lays a foundation for path planning.The front end point cloud data optimization algorithm based on voxel filtering and the map alignment algorithm based on the linear scoring system have been successfully applied in sweeping robots,effectively improving the positioning and mapping effects of mobile robots.The research on resampling strategy based on partition management can further improve the positioning and mapping accuracy of mobile robots,and has guiding significance for the promotion and application of mobile robots.