Particle filter for fault diagnosis and robust navigation of underwater robot

Abstract

A particle filter (PF)-based robust navigation with fault diagnosis (FD) is designed for an underwater robot, where 10 failure modes of sensors and thrusters are considered. The nominal underwater robot and its anomaly are described by a switching-mode hidden Markov model. By extensively running a PF on the model, the FD and robust navigation are achieved. Closed-loop full-scale experimental results show that the proposed method is robust, can diagnose faults effectively, and can provide good state estimation even in cases where multiple faults occur. Comparing with other methods, the proposed method can diagnose all faults within a single structure, it can diagnose simultaneous faults, and it is easily implemented.