Objective To address the challenges in multi-AUV formation maneuvering—such as limited state perception and transmission capabilities, acoustic communication delays, data loss, and reduced observability due to the lack of position information exchange—this study proposes an event-triggered metrology-communication unified framework with a Lyapunov-based model predictive formation control method (ETMCU-LMPC). The proposed approach aims to enhance formation stability and tracking accuracy.

Method  First, by integrating the formation communication topology with system states, an event-triggered mechanism based on state observation is established. This mechanism leverages relative measurements among AUVs to mitigate delays and data loss caused by acoustic communication failures, while improving system observability in the absence of position information exchange. Second, a distributed model predictive controller based on Lyapunov theory is designed. The controller employs backstepping to construct contractive constraints, ensuring recursive feasibility, and incorporates adaptive Kalman filtering (AKF) to compensate for measurement noise, thereby guaranteeing closed-loop stability.

Results Simulation results for a five-AUV formation (1 leader and 4 followers) show that, compared with the traditional LMPC, the proposed ETMCU-LMPC method reduces the convergence time from 8 s to 6 s, the maximum error from 1.12 m to 0.36 m, and the steady-state error from 0.57 m to 0.06 m. In addition, the control input exhibits improved stability.

Conclusion The proposed method effectively addresses communication anomalies, enhances the reliability of multi-AUV formations under conditions of limited state perception and transmission, and demonstrates significant practical engineering value.

Results  Simulation results of the formation control for five AUVs (1 leader and 4 followers) show that, compared with the traditional LMPC, the proposed ETMCU-LMPC method reduces the convergence time from 8 s to 6 s, the maximum error from 1.12 m to 0.36 m, and the steady-state error from 0.57 m to 0.06 m. Additionally, the control input exhibits greater stability.

Conclusion The proposed method can effectively cope with communication anomalies, improve the reliability of multi-AUV formations under scenarios with limited state perception and transmission, and thus possesses practical engineering significance.