论文部分内容阅读
传感器的可靠性及准确性对制冷机系统的可靠控制和系统的最优运行起着至关重要的作用。同时,传感器的读数也是进行部件故障诊断的基础。本文提出了基于主元分析法的制冷机传感器故障诊断方法,该方法的主元分析模型由离心式制冷机系统中的相关测量变量在正常条件下的观测样本构成。这一方法利用这些变量在正常条件下的相关性来对传感器的测量观测值进行故障检测与诊断及测量重构,并分别用Q-统计及Q-分布图来对传感器故障进行检测及诊断。本文利用实验室离心式制冷机的试验数据对这一基于主元分析法的传感器故障诊断方法进行了验证。
The reliability and accuracy of the sensors play a crucial role in the reliable control of the chiller system and in the optimal operation of the system. At the same time, the sensor reading is also the basis for component failure diagnosis. In this paper, a fault diagnosis method of refrigerator sensor based on principal component analysis (PCA) is proposed. The principal component analysis model of the method is composed of the observed samples under the normal conditions from the relevant measured variables in the centrifugal chiller system. This method uses the correlation of these variables under normal conditions to detect, diagnose, measure and reconstruct the measured values of the sensors and detect and diagnose the sensor faults respectively using Q-statistics and Q-distribution graphs. In this paper, we use the experimental data of laboratory centrifugal chiller to verify the sensor fault diagnosis method based on principal component analysis.