Safe Persistent Coverage Control with Control Barrier Functions Based on Sparse Bayesian Learning

Abstract

In this paper, we propose a control algorithm to explore an unknown environment while guaranteeing the safety of robots by learning safety constraints from sensor information. A sparse Bayesian classifier (SBC) is trained to estimate the probability that the robots will not collide with obstacles at each point based on the local distance data to obstacles obtained from onboard sensors. Then, we propose a control barrier function (CBF), named an SBCBF, which is used to avoid obstacles estimated by the SBC. We also develop a persistent coverage control based on the SBCBF for exploring the area keeping the robot at a given safety level. Furthermore, we build an online control algorithm that integrates the SBCBF synthesis and safe persistent coverage control. Finally, we demonstrate the effectiveness of the proposed algorithm by the simulation and experiment.