Decentralized Full Coverage of Unknown Areas by Multiple Robots With Limited Visibility Sensing

Abstract

This letter addresses the full coverage problem of unknown convex and concave two-dimensional (2-D) areas by multiple robots with limited visibility sensing and communication range. The areas are initially unknown to the multiple robots, and the number of robots is not predefined. In order to accomplish the full coverage task, the robots are released iteratively and coordinate to cover the unknown areas by exchanging local information with the neighboring robots within the communication range. A novel continuous max-sum algorithm with gradient descent is proposed to provide a decentralized and efficient solution. Moreover, the proposed method can be applied to omnidirectional/fan-shaped sensor models with various sensing ranges considering the onboard sensors of robots may vary in applications. Simulation studies are carried out to validate the effectiveness and efficiency of the proposed approach to achieve full coverage of unknown convex and concave 2-D areas.