An integrated decision, control and fault detection scheme for cooperating unmanned vehicle formations
Nicolas Lechevin, Camille Alain Rabbath, Madhavan Shanmugavel, Antonios Tsourdos, Brian A. White · 2008
We propose a hierarchical and decentralized scheme for integrated decision, control and fault detection in cooperating unmanned aerial systems flying in formations and operating in adversarial environments. To handle, in a cooperative fashion, events that may adversely affect the outcome of a multi-vehicle mission, events such as actuator faults, body damage, network interruption/delays, and vehicle loss, we present a decision-control system whose architecture comprises three main components: formation control and trajectory generation, abrupt and nonabrupt fault detection, and decision-making relying on optimization under uncertainty. The scheme seeks to provide the most effective team adaptation to contingencies despite partially known environments and limited available information. The integrated decision, control and fault detection scheme is demonstrated numerically by means of high-fidelity, nonlinear 6-DOF simulations of multiple formation flying airships. For a rendezvous mission, the paper shows that concurrent nonabrupt and abrupt type faults can be detected and effectively compensated for both at the formation control and at the decision-making levels, despite network mishaps, which represents a novelty in itself.