Research on fault diagnosis for ship course control system

This paper analyzes the fault information of ship´s course control system, and establishes a fault diagnosis model based on fuzzy nerve network algorithms. Using the fuzzy logic processing data so as to make full use of the experience and knowledge, using the nerve network so as to avoid some problems of the complications fault tree diagnosis system, such as matching conflict, combination explosion, and infinite recursion. Also adopting the improved BP arithmetic to train the nerve network which can solve the problems of convergence speed and convergence surge. The fault diagnosis result shows this fault diagnosis system has strong robustness and generalization. That method uses model free diagnosis, so that is easy to learn by self and perfect system function constantly, and has some theories and engineering application value.