Research On Simultaneous Location And Mapping Algorithm Of Mobile Robot Based On Lidar

With the development of robot technology and the promotion of intelligent transformation of traditional manufacturing industry,mobile robot technology has attracted more and more attention from academia and industry.This topic comes from the school enterprise cooperation project.According to the material transportation demand in the metal processing workshop,we are aming to design a mobile robot for the material transportation in the metal processing workshop.In this paper,the kinematics of mobile robot,multi-sensor fusion algorithm and laser SLAM technology are studied.The main contents of this paper are as follows:(1)Based on the analysis of the characteristics of the traditional metal machining workshop in manufacturing industry,the design requirements of the mobile robot are clarified,the control system of the mobile robot is decomposed into functional modules;The kinematic model and the track model of the four wheel Mecanum mobile robot are deduced,which provides the research basis for the mobile robot’s initial pose estimation of simultaneous location and mapping.(2)Wheel slip and other situations will introduce errors.In order to make up for the lack of the accuracy of initial pose estimation based on the wheel odometer,the calibration method of the wheel odometer kinematic model and the least square calibration method based on the lidar are established;Based on the extended Kalman filter algorithm,the sensor information of the wheel odometer and IMU are integrated,and the position and attitude of the robot after the fusion are taken as the mobile robot’s initial pose of the SLAM system.(3)The particle filter theory,the traditional laser SLAM algorithm and the improved rbpf-slam algorithm for optimizing the proposed distribution and adaptive resampling are studied.Aiming at the problem of computing efficiency degradation caused by the large number of particles in the Gmapping algorithm,an optimized RBPF-SLAM algorithm based on adaptive particle number of KLD resampling is proposed,which can adaptively adjust the number of particles participating in resampling according to the current distribution of particle swarm and improve the efficiency of the algorithm.(4)The simulation experiment is carried out.The traditional algorithm and the improved algorithm are compared on the self-built experimental platform,and the effectiveness of the improved algorithm is verified by the field experiments in three different indoor environments.