基于聚类原理的涡扇发动机飞行剖面逐步分类方法

Stepwise Classification Method of Turbofan Aero-Engine Flight Profile Based on Clustering Principle

提出了一种针对涡扇发动机飞行剖面分类方法。该方法以某发动机的87个飞行实测起落数据为输入,选择反映涡扇发动机工作特性的飞行高度、马赫数和转速3个参数作为飞行剖面类别的判别依据,通过提取3个参数的特征值,基于聚类原理,运用逐步分类和散点图方法,将发动机的飞行剖面分成6类。与其他飞行剖面分类法相比,该方法选用的参数更全面,运用图形的方式更直观,考虑参数特征值的离散程度实现了降维,使分类具有可操作性。

A classification method for turbofan engine flight profile is proposed.Three parameters,such as flight height,Mach number and rotation speed,which reflect the working characteristics of a turbofan engine,were selected as the basis for distinguishing the flight profile categories.The characteristic values of the three parameters were studied and extracted.By using the cluster-based scatter diagram method,the 87 flight profiles of a turbofan engine were successfully divided into six categories in two steps.Compared with other flight profile classification methods,the parameters used for classification in this method are more comprehensive.By using a scatter diagram,this method makes the classification more intuitive.Considering the degree of dispersion of the parameter’s characteristic value,this method successfully reduces the dimension and makes the classification more operable.