Diagonalization of Covariance Matrix in Simultaneous Localization and Mapping of Mobile Robot

The purpose of this study is to analyze the effects of covariance state update by means of modified algorithm of diagonal matrix using eigenvalue, and diagonalization function in MATLAB on the computational cost of extended Kalman filter based Simultaneous Localization and Mapping (SLAM). The multiplications of the covariance matrix with other parameters increase its dimension, which is twice the number of landmarks and might result in erroneous estimation. This motivates this study in searching for ways to reduce the computational cost of the covariance matrix without minimizing the accuracy of the state estimation using eigenvalue method. The matrix diagonalization strategy which is applied to the covariance matrix in EKF-based SLAM must be examined to simplify the multiplication procedure. Therefore, improvement is needed to find better diagonalization method. Simulation results demonstrate that MATLAB’s built-in diagonalization function can reduce the computational cost.