Target localisation and circumnavigation by multiple nonholonomic agents using bearing-based measurements

This paper investigates the localisation and circumnavigation problem of an unknown target by multiple agents in 2D space, where the target may either be stationary or move at a bounded velocity. The agents under consideration are subject to nonholonomic constraints, and they cannot directly acquire the target's position and velocity, while they can only measure the target's bearing information. Evidently, compared to some existing literatures where agents can directly obtain the target's position, the scenario considered in this paper is more realistic and challenging. A novel three-component control algorithm is proposed to enable agents to move around the target in an equiangular spaced formation, independant of the alignment of their local coordinate systems. Specifically, an autonomous stable exosystem is first introduced to generate the circular reference trajectories. On this basis, a position estimator is used to locate the target based on bearing measurements for each agent. Concurrently with the estimation process, a control protocol is developed to simultaneously guide the agents in tracking the reference motion of the exosystem. The stability and effectiveness of the control algorithm are validated by theoretical analysis, numerical simulations and a physical experiment.