Bio-Inspired Adaptive Cooperative Control of Heterogeneous Robotic Networks

Abstract

We introduce a new adaptive cooperative control strategy for robotic networks comprised of heterogeneous members. The proposed feedback synchronization exploits an active parameter adaptation strategy as opposed to adaptive parameter estimation of adaptive control theory. Multiple heterogeneous robots or vehicles can coordinate their motions by parameter adaptation analogous to bio-genetic mutation and adaptation. In contrast with fixed gains used by consensus theory, both the tracking control and diffusive coupling gains are automatically computed based on the adaptation law, the synchronization errors, and the tracking errors of heterogeneous robots. The optimality of the proposed adaptive cooperative control is studied via inverse optimal control theory. The proposed adaptive cooperative control can be applied to any network structure. The stability proof, by using a relatively new nonlinear stability tool, contraction theory, shows globally asymptotically synchronized motion of a heterogeneous robotic network. This adaptive cooperative control can be widely applied to cooperative control of unmanned aerial vehicles (UAVs), formation flying spacecraft, and multi-robot systems. Results of the simulation show the effectiveness of the proposed adaptive cooperative control laws especially for a network comprised of heterogeneous members.