Landmark Sequence Data Association for Simultaneous Localization and Mapping of Robots

Abstract

Abstract The paper proposes landmark sequence data association for Simultaneous Localization and Mapping (SLAM) for data association problem under conditions of noise uncertainty increase. According to the space geometric information of the environment landmarks, the information correlations between the landmarks are constructed based on the graph theory. By observing the variations of the innovation covariance using the landmarks of the adjacent two steps, the problem is converted to solve the landmark TSP problem and the maximum correlation function of the landmark sequences, thus the data association of the observation landmarks is established. Finally, the experiments prove that our approach ensures the consistency of SLAM under conditions of noise uncertainty increase.