基于核近邻非参数回归的航材消耗预测

Prediction of Aviation Material Consumption Based on Nonparametric Regression of Nuclear Nearest Neighbor

摘要:

针对某型直升机服役时间短、机型较新、备件采购周期长等特点,使用核近邻非参数回归对平稳型航材建立消耗预测模型。对比多种回归方法,证明基于非参数回归的航材消耗模型对区间预测具有较好的效果。

Abstract:

According to the characteristics of a certain type of cited helicopter, such as short service time, relatively newmodel and long spare parts procurement cycle, the non-parametric regression algorithm was used to establish the aviationmaterial consumption model. Based on the classification of aviation materials, the consumption prediction model of station?ary aviation materials was established by using nonparametric regression of nuclear nearest neighbor. By comparing vari?ous regression methods, it is proved that the model based on non-parametric regression has a good effect on interval pre?diction.