Research On Mobile Robot Isimultaneous Localization And Mapping

With the diversification of the mobile robot application environment,how to explore the unknown environment has become an important issue in the research of mobile robot.When a mobile robot is operating in an unknown and complex environment,it is necessary to map the unknown environment and accurately locate its position,which helps the robot to perform further motion planning and accomplish the task better.The robot constantly locates itself by using the constructed map,which is the problem of Simultaneous Localization And Mapping(SLAM)of an intelligent mobile robot.In this paper,based on the study of advanced algorithms for SLAM problem of intelligent mobile robot at home and abroad,the paper makes an in-depth analysis and research on the algorithm,and makes some improvements to the shortcomings in the algorithm.First of all,the coordinate system of the SLAM research institute of mobile robot is established.Based on this coordinate system,relevant models of mobile robot are established,such as environment feature model,motion model and observation model.Based on these models,the classical SLAM algorithm FastSLAM2.0 based on probability is realized.The Unscented Transform is introduced into the FastSLAM algorithm to form the UFastSLAM algorithm to reduce the uncertainty of the particle filter sampling,while eliminating the linearization error caused by the Kalman filter,and UFastSLAM algorithm and EKF-SLAM and FastSLAM algorithm in The same experimental environment conducted a comparative analysis.Secondly,aiming at the problem of particle exhaustion in Particle Filter(PF)used in UFastSLAM algorithm,an adaptive diffusion particle swarm optimization algorithm is proposed based on the selection of variance and Euclidean distance Diffusion Particle Swarm Optimization(DPSO)to optimize the UFastSLAM algorithm,which can make the particles' weight distribute evenly and control the particle exhaustion problem effectively by letting each particle move towards the optimal solution and the global optimal solution of this generation.And through the random particle diffusion with adaptive characteristics to maintain the diversity of particles,to avoid falling into the local optimum.By comparing the improved algorithm with the simulation experiment,the validity of the algorithm is verified.Thirdly,a SLAM method based on rodent brain hippocampal model was realized-RatSLAM.The simulation results show the effectiveness of the proposed algorithm.In addition,we compare RatSLAM with the traditional EKF-SLAM algorithm based on probability model under the same environment.The self-correcting function of the map under the algorithm and the superiority of the algorithm in the closed environment are verified.