Motion Planning for SLAM Based on Frontier Exploration

Abstract

This paper proposes a novel motion planning approach for SLAM in out-door large unknown environment. This approach uses the frontier based exploration strategy to find frontier points, and select the best one as the destination point of the robot. Navigating to the destination point will minimize the system uncertainty and improve the efficiency of exploration. The last part of this paper compares the proposed approach with traditional random SLAM from the point of efficiency and accuracy to validate its advantage.