研究目的
To design a particle filter-based robust navigation system with fault diagnosis for an underwater robot, considering various failure modes of sensors and thrusters.
研究成果
The proposed particle filter-based algorithm for fault diagnosis and robust navigation of an underwater robot, built on a switching-mode hidden Markov model, has been successfully tested in full-scale sea trials. The results confirm the algorithm's effectiveness in diagnosing faults and providing robust state estimation under various failure modes and disturbances. The method offers a compact, easy-to-implement solution with moderate computational complexity, suitable for real-time applications.
研究不足
The study focuses on specific failure modes of sensors and thrusters in an underwater robot. The applicability of the method to other types of faults or different robotic systems is not explored. Additionally, the computational complexity of running a particle filter on a switching-mode hidden Markov model may limit real-time application in systems with higher dimensionality or more complex failure modes.
