Multi-Vehicle Coordination in an Unknown Flowfield

Abstract

Cooperative control of autonomous vehicles is a robust approach for many spatiotemporal applications, including area surveillance, target tracking and environmental sampling. Prior research has generated stable control algorithms that promote collaboration of autonomous vehicles. A persistent difficulty in creating control algorithms is accounting for effects due to flowfields such as winds or tides. In this paper we propose decentralized control algorithms that stabilize a group of autonomous vehicles to a circular formation in either an unknown, spatially uniform flowfield or an unknown, spatially uniform flowfield whose direction rotates uniformly in time. The center of the formation can be prescribed or arbitrary. Prescribing the center formation allows us to apply these algorithms to a targettracking problem. We also show how a bounded turning-rate constraint, which is present in many autonomous vehicles, can be incorporated into control algorithms for a known flowfield. The theoretical results are supported by application examples that illustrate multiple vehicles tracking an accelerating target in a circular formation.

A multi-vehicle testbed for underwater motion coordination

“…This architecture can be utilized to test a multitude of coordination algorithms, such as those developed in [1], [3], and a multitude of other algorithms.…”

“…The focus of this paper is to show the utility of the SCUUL testbed for UUV navigation, sampling, performance and control. Examples of such algorithms are shown in [4], [1], [3],and [2]. Section 2 will describe the testbed and it's capabilities in detail.…”