An Adaptive SVSF-SLAM Algorithm in Dynamic Environment for Cooperative Unmanned Vehicles

Abstract

The aim of this paper is to accomplish the Unmanned Ground Vehicle (UGV) full self-governance by developing tools that are able to give an accurate automatic localization in unknown environment. Many techniques have been developed to make the most type of sensors solving the Simultaneous Localization and Mapping (SLAM) problem in static environment; whereas, little attention has been paid to the more realistic case of a dynamic environment. A solution to SLAM in dynamic environments would open up a vast range of potential applications. Filtering strategies play an important role to solve the SLAM problem, and are used to extract knowledge of the true states typically from noisy measurements or observations made of the system. In this context, we propose the adaptive Smooth Variable Structure Filter (SVSF) based approach to solve the cooperative SLAM problem. In our work, we introduce a covariance matrix to assess the uncertainty of the adaptive SVSF and to improve its performance and increase the number of its useful applications. The proposed algorithm is validated in real-world and the obtained results confirm the efficiency in terms of computational time and robustness.