A Novel Multi-AUV Cooperative Navigation Method Using Information Incorporation

Autonomous underwater vehicle (AUV) swarms play important roles in the ocean observation and exploration. The accurate positioning is vital for the control of AUV swarms. Due to the harsh underwater environment, the positioning of AUVs is troublesome. This paper focuses on the cooperative navigation (CN) problem of the AUV swarm in a leader-slave mode and proposes a real-time deployable CN method to improve the positioning performance of slave AUVs. The slave AUV updates its positions using the positioning messages from the leader AUV with a novel adaptive extended Kalman filter (EKF) structure. Notably, the motion uncertainty of the leader AUV is incorporated in the construction of the slave AUV’s states, and a nonlinear least square (NLS) smoother is further adopted to smooth the results. Simulation results show that the proposed CN method exhibits good performances, and reduces the root-mean-square error (RMSE) by more than 70% compared with the inertial navigation system (INS).