Improved Sliding Mode Control Technique Based on the Smoothing Boundary Layer Width for Simultaneous Localization and Mapping of Unmanned Ground Vehicle

In mobile robotics, in order to execute a well-defined mission, it is necessary to recognize the detection performance in the environment by using several sensors measuring these different characteristics. These can be merged to obtain a faithful estimate, in order to definitively reduce the uncertainty of the measurement. Simultaneous Localization and Mapping (SLAM) is necessary for the process of autonomous navigation of an Unmanned Ground Vehicle (UGV) in an unknown environment without using the Global Positioning System (GPS). There are several techniques to solve the SLAM problem using the combination of low cost sensors. In our work, the UGV makes relative observations of its own motion and features in its environment. Today, the challenges in robotics focus on solving SLAM problems with uncertainty in feature positions decreasing with time for data exchange between mobile robots. This paper describes a SLAM problem based on the Smooth Variable Structure Filter (SVSF) embedded in UGV equipped with various sensors. Simulation results show the robustness of the proposed approach, which is validated in real-time and good results have been obtained with realistic conditions.