Adaptive coverage control in non-convex environments with unknown obstacles

This paper presents an adaptive control law which enables a team of mobile robots to navigate in a non-convex environment in the presence of unknown obstacles. The proposed approach relies on potential field method, adaptive control techniques and Voronoi-based decomposition. Each agent in the network tasked with starting from random initial positions and navigating to learn regions within the area with greatest importance and concentrate their coverage on these areas while avoiding collisions with the obstacles. In particular, our presented methodology is able to deal with scenarios where a priori knowledge of the search domain is not available and when the navigation region contains unknown obstacles. Simulation results are presented to validate the enhanced performance of the proposed approach.