Set-Membership Estimation Based Distributed Cooperative Localization of Multiple UAVs

This paper investigates the distributed cooperative localization problem for mobile multi-UAV systems using set-membership estimation (SME). Unlike existing SME approaches relying on prior-known anchors or beacons, we consider a fully autonomous scenario where UAVs need to collaboratively derive guaranteed positional sets using only absolute measurements and relative range measurements from neighboring agents. To handle this problem, we first propose a law-of-total-range method to rigorously quantify the estimated sets from relative measurements, which fits the general form of relative measurements between pairwise UAVs. Based on this new method, a novel distributed SME framework is developed, where each UAV can obtain the guaranteed set containing its true state with only local measurements and communications. In addition, a distributed localization algorithm based on extended constrained zonotopes (ECZ) is designed with a proposed framework. Simulation results and field experiments verify that our proposed method efficiently improves the cooperative localization accuracy and outperforms the linearization-based extended SME method.