A Nature-Inspired Distributed Robust Control Design for Ground-Aerial Vehicle Cooperation

This paper studies the problem of tracking a moving ground object that can be found in a number of applications such as surveillance, criminal and military situations. Ground vehicles can be employed for the tracking, but are greatly affected by surrounding obstacles, e.g. building blocks, roadside objects, other vehicles and passengers, etc. Hence, their cooperation with an aerial vehicle, e.g. a drone, which has a better vision from above, is proposed. This is inspired from the wolf-raven cooperative hunting behavior in nature, where the aerial vehicle acts as the raven and ground vehicles act as wolfs. However, the position and velocity sensing as well as the vision of ground and aerial vehicles are still perturbed by the surrounding environment. Therefore, a distributed robust control design is proposed for the ground-aerial vehicle cooperation to suppress the effects of the aforementioned disturbances. Conditions on the controller parameter are derived so that the proposed controller is robust in the
$H_{\infty }$
or
$H_{2}$
sense, depending on the disturbance types. Furthermore, the bounded input constraint, as often encountered in reality, is investigated, revealing that the proposed controller is input-to-state stable with environment disturbances. Numerical simulations are then carried out to illustrate the effectiveness of the proposed nature-inspired distributed robust controller design.