基于主成分分析法的变风量空调系统传感器故障诊断

在已经建立的建筑空调系统仿真器的基础上，针对变风量（VAV）空调系统的传感器故障，提出一种基于主成分分析（PCA）和法则相结合的传感器故障诊断方法。建立了PCA模型，将由传感器测量值所组成的测量空间分解为主成分和残差两个子空间，进行故障检测后再由基于法则的策略进行故障重构。仿真试验表明，该方法不仅能够准确地检测并隔离传感器故障，而且可以初步地进行故障重构，为进一步研究传感器的故障诊断提供了必要的基础。

Principal Component Analysis (PCA) Based Fault Detection Method for Sensors in Variable Air Volume (VAV) Air-Conditioning System

A fault detection method using principal component analysis (PCA) integrated with law-based strategy was presented for detecting sensor faults in variable air volume (VAV) air-conditioning system on the base of VAV system simulation. This method, which divides the measure space into two subspaces: principal subspace and residual subspace, can detect sensor faults by PCA-based strategy and reconstruct the faults by law-based strategy. The simulation results show that this method not only can detect and isolate sensor faults, but also can briefly reconstruct the faults. It gave a necessary base for further study on sensor fault detection.