Thumbnail Image

Cooperative Hybrid Control of Robotic Sensors for Perimeter Detection and Tracking

Clark, Justin Delaney
The purpose of this work is to provide experimental results of a real-world multi-vehicle coordination application, perimeter detection and tracking, not to prove optimality, convergence, stability, etc. The tools were provided to experimentally verify the hybrid system (a dynamical system composed of discrete and continuous states). The algorithm has been extensively tested in simulation and experiments. A decentralized, cooperative hybrid system was designed and implemented that allows a group of nonholonomic robots to successfully search for, detect, and track a dynamic perimeter with limited communication, while avoiding collisions and reconfiguring on-the-fly. Furthermore, an assessment of advantages and disadvantages was drawn concerning the simulators (Matlab and Gazebo) used from this testing. Finally, experimental results were promising, but further testing is needed. Future areas of research might include more realistic outdoor tests, methods to estimate the dynamic perimeter as it evolves, and a formal analysis of the hybrid system, i.e., cycling, stability, etc.