Mobile Robot Localization in Indoor Environments Using Fuzzy Adaptive Unscented Kalman Filter and Random Tree Routing Algorithm with Fast Exploration

Abstract

In the field of mobile robot navigation, challenges such as nonlinear conditions, uncertainties, and the advancement of methods have made accurate position estimation essential. This study evaluates the effectiveness of a fuzzy-based adaptive unscented Kalman filter (FAUKF) for improving the state estimation accuracy for mobile robot localization. The proposed approach leverages the FAUKF algorithm to address noise uncertainty effectively by adaptively adjusting the covariance of the measurement noise based on a defined adaptation law. The Mamdani Fuzzy Inference System (FIS) serves as an observer, enhancing the matching law and improving overall system performance. The findings of this study demonstrate that the FAUKF algorithm provides superior position estimation accuracy compared to conventional Unscented Kalman Filter (UKF) methods. Furthermore, the research introduces an innovative navigation framework for mobile robots by integrating the Random Tree Routing algorithm with Rapidly exploring Random Tree Star (RRT*) for optimal path planning in indoor environments. The RRT* integration aims to generate efficient and optimal paths while addressing safety considerations and environmental constraints. By combining the prediction and update phases of the Kalman filter, the proposed methodology effectively minimizes the propagation of uncertainty during the localization process, thereby enabling precise localization and robust path planning for designated targets. The simulation results confirmed the effectiveness of this method in maintaining constant uncertainty levels in localization over time. The proposed adaptive method enables efficient navigation in complex environments. Path planning is a critical element in robotics applications, and the RRT*-based approach presented herein offers a comprehensive solution for generating optimal and efficient paths. By providing an up-to-date perspective, this research contributes to the evolving landscape of mobile robot localization methods. The proposed method highlights the importance of utilizing adaptive algorithms and advanced path-planning techniques to enhance navigation capabilities in indoor environments.