航空学报 > 2009, Vol. 30 Issue (8): 1380-1384

递阶偏最小二乘回归在飞机研制费用预测中的应用

王礼沅,郭基联,张恒喜   

  1. 空军工程大学 工程学院
  • 收稿日期:2008-05-28 修回日期:2008-07-28 出版日期:2009-08-25 发布日期:2009-08-25
  • 通讯作者: 王礼沅

Application of Hierarchical Partial Least Squares Regression to Development Cost Prediction of Aircraft

Wang Liyuan, Guo Jilian, Zhang Hengxi   

  1. Engineering College, Air Force Engineering University
  • Received:2008-05-28 Revised:2008-07-28 Online:2009-08-25 Published:2009-08-25
  • Contact: Wang Liyuan

摘要: 分析了飞机研制费用样本数据少、费用驱动因子众多的特点,考虑到递阶偏最小二乘回归(Hi-PLS)方法在变量规模巨大情形下进行回归建模的优势,应用递阶偏最小二乘回归方法对飞机研制费用进行预测。以战斗机机体研制费用预测为例进行研究,首先对战斗机机体研制费用驱动因子进行分组,然后应用递阶偏最小二乘回归方法对分组后的费用驱动因子进行回归建立机体研制费用预测模型。实例表明,在飞机研制费用预测方面,采用递阶偏最小二乘回归方法预测更能体现研制费用与飞机性能参数之间的关系。

关键词: 飞机, 研制费用, 递阶偏最小二乘回归, 费用预测模型, 费用驱动因子

Abstract: In the analysis of the development cost of an aircraft there are usually only small samples with a large number of cost drive factors. In view of this fact and considering the advantages of hierarchical partial least squares regression (Hi-PLS) in building a regressive model in the condition of a large number of variables, an application of Hi-PLS to aircraft development cost prediction is proposed using the development cost prediction of a fighter plane fuselage as an example. After the cost drive factors of the fighter plane fuselage development cost are grouped, Hi-PLS is applied to the regress of the grouped cost drive factors, and a fuselage development cost prediction model is established. The calculated results in the example show that Hi-PLS model can reflect satisfactorily the correlation between fuselage development cost and performance parameters of a military aircraft.

Key words: aircraft, development cost, hierarchical partial least squares regression, cost prediction model, cost drive factor

中图分类号: