Energy Conservation Diagnosis Based on Neural Network and Statistical Computing

Motor systems are highly important and are critical components in industrial processes. Up to 60% of the electricity produced in the U.S. converts into other forms of energy to provide power to equipment through motor [1]. Machinery reliability and performance can be improved with early fault diagnosis and condition monitoring; therefore, the fault diagnosis system for motor has been highlighted for years. However, among these various diagnosis systems, they all focus on the reliability but not the energy issue. Based on IEA research, the energy consumption of electric motors could be reduced up to 7% with modern engineering approach; unfortunately, the abnormal energy consumption caused by motor faults did not get the reasonable attentions. This paper is presenting an approach for energy conservation diagnosis that could predict the trend of abnormal energy consumption caused by motor faults and enabled the early energy conservation work.