Design Of Service Robot Positioning And Navigation System Based On Laser Composition

Mobile service robots have become more and more widely used in education,medical care,old-age care and cleaning,which have provided great help and convenience for people.The autonomous positioning navigation algorithm of service robot has become the key research content in this field.Simultaneous Localization And Mapping--SLAM is listed among them.The SLAM algorithm is widely used in the intelligent positioning navigation of service robots,However,in the process of localization and composition,the traditional SLAM algorithm has problems such as easy degradation of particles and easy reduction of particle diversity.Based on this,this paper will focus on the positioning,composition and path planning algorithms of service robots.Based on the systematic analysis of autonomous positioning navigation positioning algorithm,mapping method and other related theories,the integrated application of laser radar ranging,the establishment of robot pose coordinate system,odometer motion coordinate system and laser radar sensing coordinate system.Considering that the particle filter SLAM dimension is too high,this paper analyzes the particle filter SLAM based on Rao-Blackwellization theory,namely RBPF-SLAM.In view of the large deviation between the pose estimation and the actual pose in the traditional RBPF-SLAM,in this paper,the latest observation data is combined into the proposed distribution,and the proposed distribution is closer to the target.In order to solve the problem of particle degradation and particle diversity reduction in RBPF-SLAM,the effective particle numberNeff will be introduced to measure the degree of particle weight degradation and determine whether to resample,besides an improved resampling strategy is proposed,the medium weight is retained and the opportunity to copy a few low-weight particles improves the efficiency of the resampling algorithm;At the same time,the traditional RBPF-SLAM and the improved RBPF-SLAM algorithm are simulated in Matlab,and the results show that the improved estimated particles The state is closer to the true value.Aiming at the path planning and obstacle avoidance problem of service robots,this paper mainly uses A*and DWA(dynamic window method)as the research algorithms of global path planning and local path respectively,and the two path planning algorithms are simulated in the visualization tool Rviz.Furthermore,the navigation advantages of A*algorithm and DWA algorithm are combined to obtain a hybrid path planning algorithm,and implements simulation test in Gazebo simulation environment to obtain a smooth and accurate target route,and the robot can reach the destination safely.Finally,based on the MT-UROBOT robot platform with RPLIDAR,the positioning and composition experiments of the traditional RBPF-SLAM algorithm and the improved RBPF-SLAM algorithm are carried out in the experimental environment,which proves that the improved algorithm has higher positioning and mapping accuracy.At the same time,the robot path planning and obstacle avoidance experiment are completed in the experimental environment.The result shows that the navigation effect of the global path A*algorithm,DWA algorithm and its fusion algorithm reaches the expected index.