基于PCA的空调系统传感器故障诊断

针对定风量(CAV)空调系统传感器的特点,提出了一种基于主成分分析(PCA)的传感器故障诊断方法.该方法根据系统中的能量平衡关系分析焓值,以建立PCA模型;通过计算系统的SPE值、故障重构值及其统计特性,对传感器的故障进行检测、辨识和恢复.采集天津博物馆中的传感器数据,对建立的PCA模型进行传感器故障诊断和故障恢复能力的验证,对温度与湿度传感器的偏差、漂移、完全故障与准确度等级下降故障进行了仿真,结果表明这种方法对定风量空调系统的传感器故障具有很好的诊断效果、识别能力和恢复能力.

Fault Diagnosis of Sensors in Air-Conditioning System Based on PCA Method

In the view of sensors in constant air-volume (CAV) air-conditioning system, a fault diagnosis method based on principal component analysis (PCA) is proposed. PCA model is established according to energy balance and enthalpy analysis in the system. The sensor faults are identified and recovered by calculating square prediction error (SPE) index,reconstruction index and its statistic features. Using the actual data from Tianjin Museum building control system, the PCA model is proved efficient enough to detect and recover the bias, drift, complete and accuracy decrease of sensors in CAV air-conditioning system. The results show that this accuracy method has good capability in diagnosis, recognizability and recovery.