论文部分内容阅读
排气温度(EGT)是反映民用航空发动机运行状态的重要性能参数之一,本文基于发动机各个性能参数存在相关性研究的基础上,针对小样本、贫信息量、多自变量影响的情况,结合偏最小二乘回归分析,提出了一种小样本条件下考虑发动机众多因素对EGT影响的发动机性能短期预测模型。作为试验验证,文中选取了某民用飞机发动机实际飞行数据对模型进行检验,并与大样本、足信息条件下建立的多元线性回归预测模型进行对比,结果表明该模型在贫信息下也可以达到很高的预测精度,为少数据条件下发动机性能参数短期预测提供了一个有效途径。
Exhaust gas temperature (EGT) is one of the important performance parameters reflecting the operating status of civil aero-engines. Based on the correlation of each performance parameter of the engine and the influence of small sample, poor information and multiple independent variables, Partial least-squares regression analysis, a short-term prediction model of engine performance considering the influence of many factors of engine on EGT was proposed. As a test verification, this paper selects the actual flight data of a civil aircraft engine to test the model, and compares it with the multivariate linear regression model established under the condition of large sample and foot information. The results show that the model can reach very poor information High prediction accuracy provides an effective way for short-term prediction of engine performance parameters under few data conditions.