基于CBR-BP方法的油料保障需求预测方法研究  

A Study of CBR-BP Based Fuel Supply Demand Forecast Method

作　　者：秦望龙 张海越 韩戈白[1] 闫龙 王敏[1] Qin Wanglong;Zhang Haiyue;Han Gebai;Yan Long;Wang Min(The 28th Research Institute of China Electronics Technology Group Corporation,Nanjing 210007,China;Central Theater Air Force Protection Department,Beijing 100010,China)

机构地区：[1]中国电子科技集团公司第28研究所,南京210007 [2]中部战区空军保障部,北京100010

出　　处：《信息化研究》2021年第2期34-37,共4页INFORMATIZATION RESEARCH

摘　　要：针对军事行动中的后勤油料精确化保障问题,文章提出了一种基于案例推理-BP神经网络(CBR-BP)方法的油料需求精确预测方法。首先采用案例推理方法从案例库选择与待求解案例匹配度较高的油料保障案例作为神经网络的训练样本,然后基于选择的样本数据采用BP神经网络方法构建油料保障需求模型,并对待求解案例的油料保障需求进行预测。通过仿真实验对文中方法的可行性进行了验证,仿真结果表明,文中方法对于油料保障需求预测有较好的计算精度。A CBR-BP based method is proposed in order to accurately forecast the fuel support demand in military operations. In this method, a case-based reasoning(CBR) method is firstly adopted to select similar cases of the target case as the training samples. Then with the BP neural network method, a fuel support demand model is built which is used to forecast the fuel support demand of the target case. A simulation experiment is conducted and the result shows that the CBR-BP based fuel support demand forecast method can obtain the fuel support demand accurately and efficiently.

关 键 词：油料保障 需求预测 案例推理 神经网络 

分 类 号：E239[军事—军事理论]