面向批量数据处理的涡轴发动机外场状态监控方法设计与应用

Design and Application on the Method of Turbo-shaft Engine Outfield Condition Monitoring Based on Batch Data

航空发动机外场数据积累量日益攀升,亟需充分利用大量外场飞行数据来快速准确判断发动机健康状态。本文以航空涡轴发动机为对象,根据使用维护需求,开展了面向批量数据的状态监控方法研究。在发动机参数预处理基础上,提出了多进程并行计算方法进行数据解析运算,构建了基于规则的专家系统,并结合典型发动机历史数据,采用数据挖掘和统计方法设计逻辑判据,最后开发了某型状态监控软件系统并在外场试用、验证,提升了面向机群大量数据的批量化、自动化分析能力。

With the increasing accumulation of aero-engine outfield data,it is urgent to make full use of large number of outfield flight data to quickly and accurately judge the health status of the engine.In this paper,the condition monitoring method for batch data is studied according to the requirements of operation and maintenance.Based on the pretreatment of engine parameters,a multi-process parallel computing method is proposed to analyze data,and a rule-based expert system is constructed.Combined with typical engine historical data,logical criteria are designed by data mining and statistical methods.Finally,a certain condition monitoring software system is developed and tested in the field,which improves the batch and automatic analysis ability of a large number of data for the cluster.