研究目的
To design a particle ?lter (PF)-based robust navigation with fault diagnosis (FD) for an underwater robot, considering 10 failure modes of sensors and thrusters, and to achieve FD and robust navigation through a switching-mode hidden Markov model.
研究成果
The experimental results confirm that the performance of the fault diagnosis was generally good and that the proposed algorithm provided robust and efficient state estimation for the ROV under different combinations of failure modes, signal artifacts, and disturbances.
研究不足
The study focuses on a specific underwater robot (ROV Minerva) and its navigation system, which may limit the generalizability of the findings to other underwater vehicles or systems. Additionally, the computational complexity of the PF algorithm and the need for real-time implementation may pose challenges in certain applications.
