基于MSPCA的液体火箭发动机试车台氢供应系统传感器故障诊断方法

研究了多尺度主元分析方法在传感器故障诊断中的应用问题。利用小波变换得到传感器信号在各个尺度上的系数，然后根据尺度系数矩阵建立主元分析模型进行传感器故障诊断。设计了固定窗长的移动窗口，对窗内数据进行小波变换后，用最后一个尺度系数计算平方预报误差进行故障检测，采用传感器有效指标这种具有定量辨识标准的参数对故障传感器进行辨识。最后，以液体火箭发动机试车台氢供应系统几个关键传感器的故障诊断为例验证了这种方法实用性和有效性。

Method of Sensor Fault Diagnosis in Liquid Rocket Engine Ground Testing Bed Hydrogen Providing System Based on MSPCA

Multiscale principal component analysis for sensor fault diagnosis was studied in this paper. Wavelet transform was utilized to get coefficients of each scale. Based the coefficient matrix, principal component analysis model was established to diagnose sensor fault. A moving window with constant length was designed in application. After wavelet transform of the data from the window, square prediction error was calculated with the latest coefficient to detect failure. Meanwhile, sensor validate index with quantitative standard was calculated to identify the faulty sensor. Finally, the applicability and effectiveness of the proposed methodology is illustrated by some sensors in the rocket engine ground testing bed Liquid hydrogen providing system.