Item

Abstract

This paper presents a novel distributed control strategy that enables multi-target exploration while ensuring a time-varying connected topology in both 2D and 3D cluttered environments. Flexible continuous connectivity is guaranteed by gradient descent on a monotonic potential function applied on the algebraic connectivity (or Fiedler eigenvalue) of a generalized interaction graph. Limited range, line-of-sight visibility, and collision avoidance are taken into account simultaneously by weighting of the graph Laplacian. Completeness of the multi-target visiting algorithm is guaranteed by using a decentralized adaptive leader selection strategy and a suitable scaling of the exploration force based on the direction alignment between exploration and connectivity force and the traveling efficiency of the current leader. Extensive MonteCarlo simulations with a group of several quadrotor UAVs show the practicability, scalability and effectiveness of the proposed method.