ABSTRACT
Summary
A ship propulsion system is a mechatronic system. Online monitoring and fault diagnosis are crucial to achieving high performance levels from such a system. In this chapter, an online monitoring and fault diagnosis system for torsional vibration in the shafting of a ship propulsion system is presented. First, the typical fault types of ship shafting are outlined. Second, the fault diagnosis principles based on the back-propagation neural network and the Kohonen self-organization neural network are presented. Third, a simulated test bed is developed, and its architecture and functions are introduced. In conclusion, experimental results are presented and discussed to validate the developed approaches.
