基于SKPCA的卫星整流罩空调系统传感器故障检测研究

摘要

Aiming at higher reliability requirement for sensor fault detection for air conditioning system of satellite fairing,on the basis of analysis on common fault modes,which focuses on improving the performances of traditional detection methods based on data-driven,adopt a fault detection method based on statistics kernel principal component analysis( SKPCA ),sensor fault detection model based on SKPCA is established,in addition,make validation experiment for common bias faults. Correctness and effectiveness of SKPCA on fault detection is verified by results of experiment.%针对卫星整流罩空调系统对传感器故障检测的高可靠性的要求，在分析常见故障模式的基础上，采用了基于统计量核主元分析（ SKPCA）的故障检测方法，建立整流罩空调系统传感器故障检测模型，对整流罩空调系统传感器容易出现的偏置故障进行了验证实验。实验结果验证了SKPCA方法在整流罩空调系统传感器故障检测中的正确性和有效性。