基于粗糙集和神经网络的舰载直升机航材消耗预测研究

Demand Prediction of Carried Aviation Materials for Ship-borne Helicopters

针对舰载直升机航材消耗影响因素的非定量性以及冗余性等问题,以舰载直升机操作平台某机械类易耗件的消耗预测为例,分析了影响其消耗的因素,利用粗糙集的理论知识约简方法去除冗余,选择出主要影响航材消耗的因素,利用神经网络建立了航材消耗的预测模型。最后比较了该模型的预测效果和直接神经网络预测的效果,显示了模型的实用性和有效性,为航材的消耗预测提供了一种有效的决策方法。

In view of the ship-borne helicopter material consumption influence factors of the quantitative and redundancy, as ship-borne helicopter operation platform, a precision instrument parts consumption forecast as an example, the factors affecting the consumption are analyzed, the theory of rough set knowledge reduction method is used to remove redundancy, the main factors af fecting the consumption of material are chosen, material consumption is established based on the neural network prediction model. Finally the prediction resuhs of the model and the effect of neural network prediction are compared, the practicability and validity of the model are shown, an effective decision is provided for material consumption forecast method.