A novel expert system of fault diagnosis based on vibration for rotating machinery

To avoid significant losses of rotating machinery which works in high-speed and heavy load for long-term, it is necessary to find faults by means of vibrations. A novel expert system of vibration fault diagnosis based on artificial intelligence for rotating machinery was presented, in which a equipment property database is established to obtain the symptom frequencies of fault of components, such as rotor, roll bearing and gear box, of equipment, so any fault can be found quickly and effectively and then the losses of fault can be reduced and further eliminated. The diagnostic reasoning engine of the system combined the forward reasons method and the forward-backward hybrid reasons method. It is proved by the diagnostic examples that the system is reasonable and scientific in structure, quick and reliable in diagnosis.