On the attitude consensus of rigid bodies using exponential coordinates

Abstract

Consensus is an emergent behavior that occurs when the robots or agents of a robotic network reach an agreement and converge to a common state. Consensus is essential in many multi-robot coordination tasks. This paper addresses the attitude consensus problem in a network of fully actuated rigid bodies. We propose two consensus control laws based on the exponential coordinates of rotation. The first controller uses absolute attitude measurements, whereas the second one employs relative measurements. Both control algorithms avoid potential singularities of the exponential coordinates and achieve almost global asymptotic stability. Under certain conditions on the network topology, the proposed consensus protocols work for either bidirectional and unidirectional communication channels. Finally, we provide numerical simulations and experimental results to assess the performance of the proposed control laws.