针对主成分分析法检测敏感度不足的问题,文中提出基于稀疏主成分的空调传感器故障诊断方法.该方法使用弹性网稀疏化载荷矩阵,减少主元与变量的关联,增强主成分的可解释性,从而提高故障检测的敏感度.某商场空调系统传感器的实验仿真结果证明,文中方法较主成分方法具有较高的准确率并且提高了检测的敏感度,具有很高的应用价值.%Aiming to problem of the lack of sensitivity of principal component analysis ( PCA ) ,a sensor fault diagnosis method based on spare principal component analysis ( SPCA)was proposed .This method applied elastic net to thin the load ma -trixes, reducing the relevancy between principal component (PC) and multiple variables, which improved PC's interpretability and enhanced the sensitivity of failure detection. The experimental results show that comparing with PCA method , the proposed method achieves high accuracy of fault diagnosis and enhances the sensitivity of failure detection of sensor in air -conditioning sys-tem.The proposed method has application value .
展开▼