一种基于有偏估计的飞机备件需求预测模型  

Aircraft spare parts demand forecasting model based on biased estimation

作　　者：刘登一 侯胜利[2] LIU Dengyi;HOU Shengli(Graduate Brigade of Air Force Logistics Academy,Xuzhou 221000,China;Department of Air Materials and Four Stations of Air Force Logistics Academy, Xuzhou 221000,China)

机构地区：[1]空军勤务学院研究生大队,江苏徐州221000 [2]空军勤务学院航材四站系,江苏徐州221000

出　　处：《兵器装备工程学报》2022年第4期152-157,共6页Journal of Ordnance Equipment Engineering

摘　　要：以提高某种具有多重共线性需求规律的飞机备件,需求预测的快速性和准确性为目的,在综合考虑各影响因素关联性的基础上,提出了一种基于有偏估计的飞机备件预测模型。该模型针对某特定器材进行多重共线性判别,利用主成分分析法对特征变量进行降维,利用以岭回归为代表的有偏估计方法对飞机备件进行需求预测。通过某型飞机备件需求的实例验证,表明该模型对具有多重共线性需求规律的备件具有较好的预测性,为飞机备件需求预测提供了一种快速有效的模型。In order to improve the rapidity and accuracy of spare parts demand forecasting for a certain aircraft spare parts with multicollinearity,this paper proposed a biased estimation based spare parts forecasting model considering the correlation of various influencing factors.The model made multicollinearity discrimination for a specific equipment and then used principal component analysis to reduce the dimension of characteristic variables.Finally,it used the biased estimation method represented by ridge regression to predict the demand of aircraft spare parts.An example of a certain type of aircraft spare parts demand shows that the model has good predictability for the spare parts with multicollinearity demand law and provides a fast and effective model for the prediction of aircraft spare parts demand.

关 键 词：有偏估计 岭回归 主成分分析 备件 需求预测 

分 类 号：O141.4[理学—数学]