Research On Pose Estimation Algorithm Of Mobile Robot Based On Entropy

With the rapid development of economy and the continuous progress of science and technology,mobile robots are playing an increasingly important role in manufacture and people's daily,and the pose estimation of mobile robots is a key problem that must be solved to realize the application of robots.This thesis focuses on the pose estimation algorithm of mobile robot under non-Gaussian noise to improve the accuracy of pose estimation of mobile robot.In the simultaneous localization and mapping(SLAM)framework,the research focuses on the back end optimization algorithm and the front end set registration algorithm,which affect the accuracy of robot pose estimation.To address the problem of low pose estimation accuracy of traditional filtering algorithm for mobile robots in non-Gaussian noises,we proposed an pose estimation algorithm based on the combination of iterative unscented Kalman filter(IUKF)and maximum correntropy(MC),named as MC-IUKF.In this thesis,MC is used to deal with non-Gaussian noises,a cost function based on MC is constructed,and then the Levenberg-Marguardt method(LM)is used to optimize the cost function.Then,the iterative updating process of state and covariance is derived,and the updating steps of IUKF are improved.The simulation shows that the proposed algorithm has better estimation accuracy than the traditional filtering algorithm in the non-Gaussian noise environments,and it has favorable stability.In three-dimensional space,the accuracy of point cloud registration directly affects the accuracy of robot pose estimation.The research on point cloud registration algorithm is an important link to improve the accuracy of robot pose estimation.To solve the problem of low registration precision of Iterative Closest Point(ICP)algorithm when dealing with Point cloud data with noise or outliers,we proposed a point set registration algorithm(MC-ICP)based on the combination of MC and ICP.In this thesis,correntropy is first applied to the point set registration problem,and then a point set registration function based on correntropy is proposed.Then,the pose conversion parameters between point sets are optimized.The proposed algorithm can handle the point set registration problem with noises and outliers well.Finally,the proposed algorithm is applied to the pose estimation of mobile robots.Simulation shows that compared with the traditional ICP algorithm,the MC-ICP algorithm has higher registration accuracy and better robustness,and also effectively improves the pose estimation accuracy of the mobile robot.The research in this paper is an important part of the national natural science foundation project "collaborative positioning and mapping of heterogeneous unmanned systems in satellite signal refusing environment".The research provides insight into the pose estimation of mobile robots,and is of theoretical and practical significance for promoting the popularization and application of robots.