基于n阶灰色补偿因子的装备维修保障费用预测模型  

Prediction model of equipment maintenance support cost based on n-order gray system theory

作　　者：单黎黎[1,2] 何向东 钟志明[2] 袁继军 李梦璇 

机构地区：[1]解放军理工大学军事仿真教研室 [2]73686部队

出　　处：《江苏大学学报（自然科学版）》2014年第6期685-692,共8页Journal of Jiangsu University：Natural Science Edition

基　　金：国家自然科学基金资助项目(70971137)

摘　　要：为了全面提高装备维修保障费用的合理使用度,对装备维修保障费用组成进行了分析,构建了装备维修保障费用结构模型;对装备维修保障数据进行了累加生成,将x轴向下平移处理以选择合适的预测模型;为了提高普通GM(1,1)模型的预测精度,提出了一种基于n阶补偿因子的GM(1,1)预测模型,并给出寻优算法以求补偿过程中的最优值,同时通过该算法弱化了"过补偿"现象.通过试验比较了普通GM(1,1)模型和BP神经网络模型.结果表明,所提出算法和模型相对于其他预测算法具有较高的可靠性和一致性.To improve the using rationality of equipment maintenance support（EMS） cost,the concept of ″compensation factor″ and a new GM（1,1） model were proposed based on the n-order compensation factor.A calculation flow was designed to obtain the ideal value of the proposed model,which could also weaken the phenomenon of ″over-compensation″ in the forecasting process.The actual data of EMS cost was analyzed,and the panning transformation as well as accumulative generation was realized to achieve the actual data sequence because the actual data of EMS cost have a strongly monotonic increasing regularity.The prediction model of EMS cost was constructed to solve the solution by the proposed method,the calculation flow and the processed data.The experiment was conducted to compare the introduced model with ordinary GM（1,1） and BP neural network.The results show that the prediction accuracy of the proposed model is higher than those of the other two methods,which indicates that the prediction model based on n-order compensation factor is useful and accurate.

关 键 词：装备维修保障 费用 预测 灰色系统理论 n阶补偿因子 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]