Answers are grounded in this paper and cite their sources.
Swarms of biomimetic underwater robots offer potential for marine environmental monitoring and underwater exploration through coordinated operation. Accurate and reliable navigation is essential for state sharing and coordinated motion. However, cooperative navigation is challenged by dynamic membership, temporally correlated ranging errors, and unknown correlations among node estimates. This study proposes a Kullback–Leibler averaging fractional-order cubature filter based on the labeled multi-Bernoulli framework (KLFC-LMB). The algorithm represents the time-varying set of cooperating autonomous underwater vehicles (AUVs) within the labeled multi-Bernoulli (LMB) framework, using existence probabilities and conditional state densities to jointly characterize membership uncertainty and member motion states. A fractional-order colored-noise compensation term is constructed from historical ranging residuals to account for temporal correlations in ranging errors. Before each relative measurement update, Kullback–Leibler averaging (KLA) combines the state predictions maintained by the local and neighboring AUVs for the same labeled AUV to construct a fused prior. Measurements from neighboring AUVs are processed sequentially at each time step to obtain the local AUV’s state estimate and the local LMB posterior over cooperating members. Simulations and a surface-vessel field evaluation show that the proposed method maintains stable member-set estimation and cooperative localization under changing swarm membership and temporally correlated measurement noise. In the sea trial, KLFC-LMB reduced the position RMSE and OSPA distance by 24.24% and 67.40%, respectively.
No takes yet. Share an insight, caveat, or question.
Xing et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: