FLC tuned with Gravitational Search Algorithm for Nonlinear Pose Filter

Abstract: Nonlinear pose (\textit{i.e,} attitude and position) filters are characterized with simpler structure and better tracking performance in comparison with other methods of pose estimation. A critical factor when designing a nonlinear pose filter is the selection of the error function. Conventional design of nonlinear pose filter design trade-off between fast adaptation and robustness. This paper introduces a new practical approach based on fuzzy rules for on-line continuous tuning of the nonlinear pose filter. Each of input and output membership functions are optimally tuned using graphical search algorithm optimization considering both pose error and its rate of change. The proposed approach is characterized with high adaptation features and strong level of robustness. Therefore, the proposed approach results of robust and fast convergence properties. The simulation results show the effectiveness of the proposed approach considering uncertain measurements and large error in initialization.