Research Of Mobile Robot Simultaneous Localization And Mapping Based On Rao-blackwellized Particle Filter

As the important foundation of intelligent mobile robot navigation, the SimultaneousLocalization and Mapping problem of mobile robot gets great attention from manyscholars. Simultaneous Localization and Mapping means that the robot orientate itsown position and at the same time create the map, the localization according to themap’s precision and the map according to the localization’s precision. As theimportant foundation and key technology of intelligent mobile robot navigation, SLAMproblem gradually become a hot topic in the field of mobile robot in the past tenyears.EKF SLAM algorithms have some shortcomings: high computationalcomplexity, noise must be white Gaussian noise, data association problem. We useanother algorithm to realize SLAM and solve these problems; the main work is asfollows:(1) In this thesis, we introduce the basic theory of robot SLAM problem andprobability model, and based on this we build the SLAM system model, including themotion model, observation model, noise model, environment model maps and so on, andgives the sampling position, status updates, map related formula of creation, etc.(2) In this thesis, we give the SLAM algorithm based on Rao-Blackwellized particlefilter, this system is broken up into two apart: path estimation and map estimation based onpath estimation and the path estimation used particle filter algorithm as the map estimationused EKF algorithm.(3) For the importance density function, sensor information acquisition and lack ofparticles in the Rao-Blackwellized particle filter algorithm, we import the genetic algorithmand simulated annealing (SA) to the Rao-Blackwellized particle filter SLAM algorithm,and we give a Rao-Blackwellized particle filter SLAM algorithm which integrates thegenetic algorithm.