Simultaneous Localization And Mapping For Robot Based On Local Map Joining

Mobile robots'autonomous navigation depends on Robots' location.Robot localization bases on the environment map.However, within an unknown environment, in order to achieve the robot localization and map building, the robots should have the capability of Simultaneous Localization and Mapping (SLAM). Simultaneous Localization and Mapping is a process used by robots to build up a map according to the information measured by the sensors within an unknown environment, and meanwhile keeping track of their current location.Simultaneous Localization and Mapping is an important field of autonomous mobile robot research, and it is also the key to achieve autonomous navigation in unknown environment to complete tasks assigned, representing robots'perception and intelligence. EKF_SLAM method is often used, but the specific form of nonlinear function must be clear and the linear processing will generate a large number of errors and lead to the filter divergence problems.In a large environment, the complexity will increase very much.In order to solve these problems, this paper proposes a modified UKF_SLAM algorithm based on conditionally independent sub-map.This paper firstly describes the representation of the environment map, Kalman Filter and Extended Kalman Filter algorithm.Based on the establishment of the system models, the SLAM algorithm and data association algorithm are explained. Based on EKF algorithm,the basic processes of SLAM are introduced.On this basis, for solving the problems of computational complexity and the EKF linearization error in EKF_SLAM algorithms, the conditionally independent sub-map method is deeply studied.In this method, Conditionally independent sub-map is used.On the one hand, this can reduce the computational complexity. And on the other and because the sub-maps are conditionally independent,the observation information in the sub-maps can be shared, overcoming the error caused by independent sub-maps. This EKF algorithm errors caused by linearization can be solved by introducing the UKF algorithm which using some typical points to approximate the probability density of the nonlinear function. So it is effective to integrate the conditionally independent sub-map method and UKF algorithms for solving the EKF_SLAM computational complexity and linearity error. Simulation results validate this algorithm, verifing the feasibility and effectiveness of the proposed algorithm.The conclusion and future directions of research work are discussed in the last part of this paper.