Research On Particle Filter Location And Map Construction Method For Mobile Robot

In the field of robots,simultaneous localization and mapping(SLAM)is a hot and difficult problem in the field of robotics,which has attracted the attention of many researchers.Based on the study of SLAM algorithm at home and abroad,this paper makes a thorough analysis and Research on SLAM algorithm,and improves the shortcomings of the algorithm,which improves the estimation accuracy and execution efficiency of robot positioning and map building.Specific research contents are as follows:Firstly,the basic model of SLAM problem is described and the robot coordinate system involved is defined.On this basis,the motion model and perception model of mobile robot are established,and the model is used to lay a good theoretical foundation for the follow-up research work.Secondly,the particle filter based on Bayesian theory is emphatically studied,the existing problems and solutions of particle filter are expounded,and then the location algorithm of particle filter is deduced.Finally,the simulation results show that the performance of particle filter localization algorithm in dealing with non-linear problems is better than other algorithms.Then,on the basis of particle filter,aiming at the problem that the location accuracy of mobile robots decreases due to particle degradation and particle depletion in FastSLAM algorithm,this paper introduces the idea of firefly optimization into FastSLAM algorithm.The improved Firefly algorithm(FA)is used to optimize the particle sampling process,which makes the particle set tend to Mobile Robots before resampling.The real pose improves the SLAM performance of mobile robot.In the environment of landmark feature map,FastSLAM algorithm and PSO-FastSLAM algorithm FA-FastSLAM algorithm are compared in estimating robot pose and landmark features,and the effectiveness of the algorithm is verified.Finally,because the grid map does not need to define road signs in advance and is easy to create and maintain,this paper describes the FastSLAM algorithm based on grid map,introduces the proposed distribution of fusing the latest observation information,and extends the improved firefly algorithm to further adjust and optimize the particle set,so that it can be distributed near the real pose of mobile robots more quickly and slowly at the same time.The phenomenon of particle degeneration is solved.Finally,the feasibility and validity of FA-FastSLAM algorithm are verified by SLAM simulation.