Range based target localization using a single mobile robot or multiple cooperative mobile robots

Abstract

In the paper, the target localization problem is studied based on only range measurements using a single mobile robot or multiple cooperative mobile robots. Differentiating from most existing work, our work assumes that the robots have no access to global knowledge (e.g., global coordinate systems, common sense of directions, absolute positions of themselves), but it is expected that they could estimate the relative coordinate of the target in the local coordinate frame attached to themselves. This is much more useful in applications of source seeking (e.g. a radio source emitting radio waves) and local motion coordination in absence of GPS signals and reference beacons. The self-awareness motion is taken into account for the estimation based on a particle filter in order to eliminate the ambiguity caused by indefinite inertial frames. Multi-robot cooperative localization exhibits improvements of estimation accuracy and robustness in response to measurement failures. Simulation results are provided to demonstrate the effectiveness of the proposed algorithms.