Cooperative Relative Localization Optimization Method Based on Joint Constraints and Backtracking Search

In order to tackle the problem of relative localization drift during Unmanned Aerial Vehicle (UAV) swarm formation flight, we propose a collaborative optimization method that integrates joint constraints and a backtracking search algorithm. By introducing Ultra Wideband (UWB) based relative ranging, we construct a graph model for the UAV swarm and design joint constraints to optimize the relative localization. The joint constraints include relative position constraint, feasibility constraint, and global position constraint. The backtracking search method is designed to optimize these constraints, which can utilize historical estimation information to prevent the search from getting stuck in local optima. By utilizing the results from the search process, optimized relative localization can be generated. Our method is a general optimization approach applicable to various relative localization methods, including vision-based, LiDAR-based, and inertial odometry-based methods, among others. We design several simulation experiments to validate the proposed algorithm and compare it with other relative localization optimization methods. The experimental and comparative results prove the effectiveness of the proposed method.