Investigation of Sensor Fault Diagnosis in Air Handling Units Based on Wavelet Energy Entropy
This paper presents a novel fault diagnosis method for sensors in air-handling units based on wavelet energy entropy.Instead of directly comparing the numerous data under noise conditions,the wavelet energy entropy deviation is used for the fault detection and diagnosis.The actual Three-level wavelet analysis is used to decompose the measurement data captured from sensors first and then the concept of Shannon entropy is referred to define the wavelet energy entropy.Once the wavelet energy entropy is obtained,whether the sensors are faulty can be confirmed through comparing the deviation of the wavelet energy entropy residual of the measured signal and the estimated one to the preset threshold.Testing results show that the wavelet energy entropy is a sensitive indictor to diagnose the sensor faults.The deviations of wavelet energy entropy of sensors under fault-free conditions and faulty ones all exceed the threshold.The severer the fault is,the larger the residuals of the wavelet energy entropy will be.The results prove that the proposed method is valid and effective for the fault detection and diagnosis of the sensors.
Fault diagnosis Sensor Air-handling unit Wavelet energy entropy
Su-ping LI Yao-ling FAN
Department of Information Engineering, North China University of Water Resources and Electric Power, Zhengzhou 450011, China
国际会议
太原
英文
316-319
2013-01-13（万方平台首次上网日期，不代表论文的发表时间）