Adaptive Neuro-Fuzzy Extended Kaiman Filtering for robot localization

Abstract

Extended Kalman Filter (EKF) has been a popular approach in localization of a mobile robot. However, the performance of the EKF and the quality of the estimation depends on the correct a priori knowledge of process and measurement noise covariance matrices (Q <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k</sub> and R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">K</sub> , respectively). Imprecise knowledge of these statistics can cause significant degradation in performance. In this paper, the Adaptive Neuro-Fuzzy Inference System (ANFIS) supervises the performance of the EKF with adjusting the matrix Q <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k</sub> and R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">K</sub> . The ANFIS is trained using the steepest gradient descent (SD) to minimize the differences between the outputs of ANFIS and desired outputs. The simulation results show the effectiveness of the proposed algorithm.