Chaotic Initialization Particle Filter AUV Cluster Position Calibration Algorithm Based on Intragroup Distance Measurement Under Large Initial Position Error

Abstract

With the increasing diversity and complexity of maritime mission requirements, the technology of collaborative multiautonomous underwater vehicles (AUVs) has garnered widespread attention. In this domain, the positional calibration technology of AUV clusters is an integral aspect that cannot be overlooked. Traditional leader–follower AUV cluster positional calibration models and algorithms have utilized information from either a single leader AUV or multiple leader AUVs in conjunction with a single follower AUV. However, with the expansion of the scale of follower AUVs, the availability of follower–follower AUV information increases. Consequently, this article develops a novel AUV cluster positional calibration model that leverages both the distance information between leader and follower AUVs, and the follower–follower AUV distance information. The observability of this model is analyzed, and building upon this, a chaos-initialized particle filter algorithm for AUV cluster positional calibration is proposed. Finally, experiments are conducted to compare the performance of the algorithm presented in this article with the particle filtering algorithm under different initial error conditions. The results demonstrate that the proposed algorithm exhibits stable convergence speed and calibration error at low initial errors. At high initial errors, it achieves faster convergence, lower calibration error within a finite time, and enhanced stability.