Study of machine fault diagnosis system using neural networks

This research develops a machine fault diagnosis system using neural networks and spectral analysis. Generally, it is very difficult to diagnose a machine fault by conventional methods based on mathematical models because of system complexity and the existence of nonlinear factors. In this research, a neural network is applied to the fault diagnosis of the machine. The neural network has learning and memory capability. By the learning of normal and abnormal states of the object system, a new-method with neural networks is proposed which can diagnose a fault of the machine. The proposed fault diagnosis system is based on the spectrum of vibrations or sounds obtained from the operating machine, because the time series data of vibrations or sounds are complicated and include noise. The difference between normal and abnormal data becomes clearer comparing time series data. It is suitable for the detection of the fault to utilize changes of spectral data. Using this method, it is shown that it can detect unknown fault patterns. The fault diagnosis experiments are performed on both a wood slicing machine and an electromagnetic valve. The possibility of an on-line fault diagnosis system is examined through the construction of an on-line data processing system for an electromagnetic valve and it is shown that the fault diagnosis can be performed in real time. Through these results, the effectiveness of the proposed fault diagnosis system is verified.