基于马氏距离的航空发动机健康监控方法

摘要

Aero-engine health Monitoring is the key to implement engine health management ( EHM ) . Dealing with the high-dimension computation complexity ,a method which combined principal component analysis(PCA) and Mahalanobis distance (MD) was proposed.The proposed method used PCA to reduce the dimension of monitored data ,calculated the health value of engines by MD and set the sentinel values conditioned on Weibull distribution .Numeric results showed that the proposed method played a good part in engines health monitoring and anomaly detection .%航空发动机健康监控是实现航空发动机健康管理的关键技术之一。针对航空发动机健康监控高维计算复杂度高的问题，提出了一种基于主成分分析与马氏距离结合的计算方法。首先通过主成分分析对性能监测参数进行降维，并采用马氏距离计算发动机的健康参数，最终根据威布尔分布确定健康参数的警戒值。实例计算结果表明，方法能够较好反映航空发动机健康状态，并能检测运行过程中出现的异常点。