基于卡尔曼滤波和主元分析的发动机性能衰退

Engine Performance Deterioration Based on Kalman Filter and Principle Component Analysis

摘要:

在给定测量参数基础上,通过非线性估计得到对应离散时刻的部件性能参数,利用数据插值方法扩充测量参数。利用主元分析方法对这些测量参数进行分解,提取反映性能衰退的成分,以此构造反映性能变化的统计量,假定该统计量服从F分布,对性能衰退情况做进一步分析,得到在10 000时间点附近性能衰退最为严重,已超过阈值,与前面参数估计结果相符。用部件性能衰退分析方法对发动机性能状况监控,为及时采取维护措施提供了一定的指导作用。

Abstract:

Component performance parameter of corresponding discrete time was gained by non-nonlinear parame. ter estimation on basis of given measurement. Measurement was added by interplotation. Measurement was decom. posed by PCA and component that expresses performance deterioration was extracted. Statistics of performance variation was established by this, assuming the statistic submit F distribution, furtherly analyzing performance de. terioration, and the performance deterioration near 10000th was most severe, even exceeding valve value. It was. correspond with estimated result afore-mentioned. Analysis method of component performance deterioration provid. ed a reference to monitoring engine performance and adopting maintenance timely.