Cooperative Vision-Based Multi-Vehicle Dynamic Coverage Control for Underwater Applications

This paper studies the dynamic coverage control problem for underwater applications using a fleet of cooperative submarines with vision-based cameras. A sensor model for the vision-based camera is first presented. The coverage goal, which is to collect a desired amount of satisfactory quality samples at every point in a given domain, is stated in a precise mathematical statement. A coverage error function is then defined and used as a metric for the coverage quality. A gradient-descent control law is then derived. Numerical simulations are presented to illustrate the main results. Centralized and decentralized system architectures currently under investigation by the authors for an experimental test-bed are also discussed.