基于GM-BP的武警部队油料消耗预测  

Prediction of Fuel Consumption in PAP Based on GM-BP

作　　者：王镇龙 周利坤[2] 沈艳林[2] 崔韵[1] WANG Zhenlong;ZHOU Likun;SHEN Yanlin;CUI Yun(Postgraduate Group, Logistic University of PAP, Tianjin 300309, China;Department of Logistics Support, Logistic University of PAP, Tianjin 300309, China)

机构地区：[1]武警后勤学院研究生管理大队,天津300309 [2]武警后勤学院后勤保障系,天津300309

出　　处：《兵器装备工程学报》2019年第S02期159-162,共4页Journal of Ordnance Equipment Engineering

基　　金：天津市自然科学基金项目(18JCQNJ01400);博士启动金项目(WHB201707)。

摘　　要：采用了一种基于灰色神经网络的组合预测算法,建立了武警部队油料消耗预测模型。模型先运用灰色理论对油料消耗进行预测,得到一组误差序列;再采用BP神经网络对误差序列进行训练,得到误差修正序列;最后将灰色模型预测消耗量与误差序列修正值进行加和,作为油料消耗预测结果。预测结果表明,模型的预测误差明显减小,对武警部队油料消耗预测有参考应用价值。The paper puts forward a demand forecast algorithm based on GM-BP and establishes combination forecasting model based on GM-BP to predict fuel consumption.Firstly,it forecasts fuel consumption with grey model and gets a set of deviation sequence.Then,it trains deviation sequence with BP neural network and obtain updated deviation sequence.Finally,the two forecast results are added together as the forecast result of oil consumption prediction in PAP.The result shows that the forecast model based on GM-BP has high forecast accuracy and has a reference value on oil consumption prediction in PAP.

关 键 词：武警部队 油料 消耗预测 灰色模型 BP神经网络 

分 类 号：TE89[石油与天然气工程—油气储运工程]