Distributed Affine Formation Maneuvering for Networked Marine Surface Vessels Under Radom Packet Loss: A Knowledge-Based Localization Solution

Formation maneuvering (FM) of networked marine surface vehicles (MSVs) encounters challenges in preserving system performance under unreliable inter-vehicle interactions and ensuring real-time operational flexibility in dynamic environments. To solve these issues, a knowledge-based affine formation maneuver (AFM) scheme is proposed, which is composed of a virtual leader trajectory generator (VLTG), distributed hybrid localizer (DHL) and knowledge-based event-triggered communication mechanism (KETCM). Specifically, the VLTG generates the real-time guidance trajectories for networked MSVs with intuitively defined affine transformations. Based on this, the DHL allows local vehicles to realize real-time resilient AFM trajectory observations under the distributed neighboring interactions. Notably, to mitigate performance deterioration caused by packet losses, a pseudo feedback neighbor estimator (PFNE) is integrated into the DHL, which provides accurate neighbor data estimations between successive packet receptions. Based on these components, KETCM schedules efficient communication triggers while ensuring DHL performance, which is derived from evaluations of multiple historical communication knowledge to avoid unnecessary communications under random packet losses. Compared to existing methods, the key advantage of the proposed method lies in the simultaneous achievement of real-time maneuverability, AFM performance, and communication efficiency under unreliable neighboring communications. Finally, theoretical analysis and semi-physical experiments validate the effectiveness of the proposed solution.