航空发动机试飞数据野值识别与补正算法

Research on Outliers Recognition and Correction Methods for Aero-Engine Flight Test

为了降低智能实时监控系统由于数据野值导致的虚警,需要在监视系统进行故障认定前对数据野值进行识别和补正。针对实际航空发动机试飞数据的情况,设计了基于最小二乘拟合结合莱特准则的自适应野值识别算法;采用适应于发动机试飞监视的野值补正算法,达到实时监视中野值剔除的目的。通过仿真数据和实际试飞数据的验证,表明了此方法在试飞中应用的有效性。

In order to reduce the false alarm caused by data outliers in the engine intelligent real-time monitoring system, it is necessary to identify and correct the outliers before the fault identification.The self-adaption algorithm of outliers recognition of the actual aero-engine flight test data based on least square fitting and White rules is proposed.With the practical correction approach for aero-engine data, the purpose of outliers elimination has been realized.The validity of the method is verified by the simulation data and actual flight test data.