홀트-윈터스 계절성 모델을 이용한 수송기 항공전자 수리 부속품의 수요 예측

Abstract

Today's weapon systems are becoming more and more expensive and more advanced due to the revolutionary development of advanced science and technology, aircraft avionics, engine, and aircraft repair parts. The current Air Force targets 85% of the demand forecast accuracy and requires a lot of effort in order to raise the forecast accuracy. The Holt-Winters seasonal model is proposed for repair parts forecast to increase the predictive accuracy of the repair part indicating seasonality. This paper compares the predictive models of the present Air Force repair parts with the additive seasonal model and the multiplicative seasonal model of Holt-Winters. The optimal forecasts were obtained by varying the level exponential smoothing constant, trend exponential smoothing constant, season exponential smoothing constant of the Holt - Winters seasonal model and the accuracy was compared through model evaluation index. Demand forecasts for seasonal transport aircraft avionics repair accessories show that the application of the Holt-Winters seasonal model is excellent.