摘要
经济增长的必要支撑条件之一是能源消费量的迅速增长,因此做好未来能源需求预测为能源规划提供科学依据有重要的现实意义。分数阶灰色模型和BP神经网络在能源预测问题领域应用十分广泛,但两者各自存在相对缺点。本文将两者进行加权组合,建立分数阶灰色神经网络组合模型,利用全国制造业天然气消费数据,将组合模型与分数阶灰色模型、BP神经网络进行预测结果误差比较分析,验证组合模型的有效性和精确性。再利用该组合模型对2017-2021年四川省天然气消费数据进行预测分析,结果表明,四川省天然气消费量总体呈上升趋势。
One of the necessary supporting conditions for economic growth is the rapid growth of energy consumption.Therefore,it is of great practical significance to make a good forecast of future energy demand to provide scientific basis for energy planning.The fractional order grey model and BP neural network were widely used in the field of energy forecasting.In this paper,the two is combined to establish the fractional order grey neural network combined model.Using the data of natural gas consumption in national manufacturing industry,the combined model is compared with the fractional order grey model and BP neural network to analyze the error of the prediction results,which verify that the validity and accuracy of the combined model.The combined model is used to forecast and analyze the natural gas consumption data of Sichuan province from 2017 to 2021.The results show that the overall trend of the natural gas consumption in Sichuan province is on the rise.
作者
胡宇
苏铃麟
黄子萌
唐国鑫
陈兴志
Hu Yu;Su Linglin;Huang Zimeng;Tang Guoxin;Chen Xingzhi
出处
《石油工业计算机应用》
2018年第2期60-64,2,共6页
Computer Applications Of Petroleum
基金
西南科技大学大学生创新基金项目精准资助专项(项目编号:jz18-048)
关键词
分数阶灰色模型
BP神经网络
组合预测模型
四川省天然气
Fractional order grey model
BP neural network
Combined forecast model
Natural gas in Sichuan province