On the real-time receding horizon control in harbor defense

Abstract

In this paper, we develop an receding horizon control (RHC) law for controlling the pursuers in a pursuit-evasion problem arising in a harbor defense scenario and describe its real-time implementation that we apply experimentally to a robotic testbed. Our implementation of the RHC law makes use of a min-max formulation of the underlying optimal problem that must be solved at each sample time, which is solved using the method of outer approximations in conjunction with a phase I-phase II method of feasible directions, in conjunction with a network layer that abstracts each agent. We demonstrate the effectiveness of our implementation using real-time human-computer simulations, and human-robot interaction using a physical testbed comprised of model-sized hovercraft.