Range-measurement-based localization and circumnavigation of multiple irregularly moving unknown targets

A planar Dubins-car like robot is controlled by upper limited angular velocity. There is a team of irregularly moving targets in the plane. The robot measures only the relative distances to the targets and cannot distinguish among them. A sliding mode control law is proposed that drives the root mean squared distance to a target to a pre-specified value; the mean is taken over all targets. This law is justified by a non-local convergence result: it is proved that the law ensures perfectly circumnavigating the targets at exactly the required mean distance. The convergence and performance of the control law are confirmed by computer simulations.