Research On SLAM Technology Of Mobile Robot Based On Bayesian Theory

Robotics is a subject that interacts in complex fields such as sensing and artificial intelligence.It not only involves all aspects of human beings,but also represents the leading level of a national frontier technology.In this paper,the SLAM problem of mobile robot is studied on the basis of Bayesian theory and the mathematical model is constructed in its theoretical framework,which leads to two core problems of data association and state estimation.In order to improve the efficiency of data association operation in the robot SLAM problem,a hierarchical clustering algorithm based on the Joint Compatibility Branch and Bound algorithm is proposed in this paper.By hierarchical clustering,all the observation values are placed in a large cluster,and the environment is divided into small clusters according to the environment.At the same time,the local correlation results are obtained by combining the joint compatibility and the nearest neighbor data association algorithm.Finally,the matching pair with the highest joint compatibility is selected as the final correlation result.Through simulation comparison,it is concluded that this method can significantly reduce the false correlation rate and shorten the running time,which is very suitable for the environment with multiple feature points.In order to solve the problem that the CKF algorithm increases with the increase of the feature points in the SLAM problem and the volume point is easily deviated from the expected trajectory,resulting in the large error of the state estimation,this paper proposes that the volume Calman filter is used to redetermine the sample volume point by the predictive value and the square root factor in the update stage,and the volume transformation is used to obtain the statistical characteristics of the system.Combined with the new estimated value,we can improve the state estimation and further improve the accuracy of the algorithm,so that we can achieve less distortion under highly nonlinear conditions.In order to verify the accuracy of the algorithm,we compare it in three dimensions,and we can conclude that the algorithm can greatly improve the pose accuracy of robot.In the end,the simulation and comparison between location and composition will be done based on the algorithm mentioned above.