Intelligent Method of Condition Diagnosis for Rotating Machinery Using Relative Ratio Symptom Parameter and Bayesian Network

In order to effectively identify faults of a rotating machine, a new kind of symptom parameter Relative Ratio Symptom Parameter (RRSP) is proposed in this paper. Moreover, the fault diagnosis system combined with Bayesian Network is built. In this paper, the relative ratio symptom parameter is calculated by using the vibration signals measured for the condition diagnosis, and the RRSP with high diagnosis sensitivity are selected by identification index as the input into Bayesian Network. By observing and analyzing the output that is the probability of normal state and abnormal states, the Bayesian Network built for the fault diagnosis of rotating machinery is proved to be effective by real data measured in each state of the rotating machine.