PDF下载

[1]刘真　潘文菊　刘佳　温凯　宫敬.基于聚类改进的区域天然气长期需求组合预测[J].油气储运,2023,42(11):1-12.
　LIU Zhen,PAN Wenju,LIU Jia,et al.Improved combination forecasting model for regional natural gas long-term consumption based on clustering algorithm[J].Oil & Gas Storage and Transportation,2023,42(11):1-12.

点击复制

基于聚类改进的区域天然气长期需求组合预测

《油气储运》[ISSN:1000-8241/CN:13-1093/TE] 卷: 42 期数: 2023年11期页码: 1-12 栏目: 出版日期: 2023-11-25

Title:: Improved combination forecasting model for regional natural gas long-term consumption based on clustering algorithm

作者:: 刘真　潘文菊　刘佳　温凯　宫敬

Author(s):: LIU Zhen; PAN Wenju; LIU Jia; WEN Kai; GONG Jing

关键词:: 天然气; 需求预测; 聚类改进; BP神经网络; 灰色预测

Keywords:: natural gas; demand forecast; clustering improvement; BP neural networks; grey model

分类号:: TE832

文献标志码:: A

摘要:: 不同区域影响天然气需求量的因素存在差异，数据集包含的数据特征不尽相同，同时天然气长期需求预测存在样本数据少的问题，因此较难构建各区域通用的预测模型。为此，利用K-means聚类算法以2011—2020年山东省11个地级市的区域历史天然气需求量、宏观经济数据等特征的相似性对区域进行分类，根据聚类结果分类组合灰色模型/BP神经网络模型进行地区天然气需求量的预测，并与单一模型预测结果进行对比，结果表明：各城市天然气需求量预测分别使用灰色预测及BP神经网络两种模型，平均绝对百分比误差（MAPE）分别为11%、6%，BP神经网络的总体效果优于灰色预测模型；聚类筛选后重新训练的BP神经网络预测结果准确性提高至4.5%。研究成果可更加准确高效地预测中国各城市未来几年的天然气需求量，为政府和企业开展天然气资源配置计划决策提供参考建议。

Abstract:: Natural gas consumption in different regions is affected by different factors. Natural gas long-term consumption forecast meets the problem of few sample data quantity and different data features. Therefore it is impossible to build a common prediction model that can be used in various regions. In order to improve the accuracy of the prediction model, it is necessary to analyze predicatively and adopt the appropriate model. The K-means clustering algorithm is used to classify the similarity of historical natural gas consumption, macroeconomic data, population and other characteristics of each region. According to the clustering results, to construct an improved gray forecast model and BP neural network to accurately predict the regional natural gas consumption. The results show that the accuracy of the re-trained BP neural network after the K-means clustering is improved. Due to the different energy structure, economic development and regional policies in different regions, grey model and BP neural network models are used for the prediction of urban natural gas consumption in Shandong Province from 2011 to 2020 with the MAPE(Mean Absolute Percentage Error) less than 6%, and the overall effect of BP neural network is better than grey model.

相似文献/References:

[1]王勇,王晓霖,张松伟.基于HPPP法的Pigsar天然气流量计检定站工艺[J].油气储运,2015,34(9):988.[doi:10.6047/j.issn.1000-8241.2015.09.014]
　WANG Yong,WANG Xiaolin,ZHANG Songwei.HPPP-based Pigsar gas flowmeter calibration facility[J].Oil & Gas Storage and Transportation,2015,34(11):988.[doi:10.6047/j.issn.1000-8241.2015.09.014]
[2]易冲冲,王文,卢超.隐式拟合在天然气热力性质计算中的应用[J].油气储运,2014,33(3):283.[doi:10.6047/j.issn.1000-8241.2014.03.013]
　YI Chongchong,WANG Wen,LU Chao.Application of implicit curve-fitting in the calculation of thermodynamic property of natural gas[J].Oil & Gas Storage and Transportation,2014,33(11):283.[doi:10.6047/j.issn.1000-8241.2014.03.013]
[3]闫文灿,徐明.移动式天然气流量标准装置的研制与应用[J].油气储运,2014,33(4):418.[doi:10.6047/j.issn.1000-8241.2014.04.016]
　YAN Wencan,XU Ming.Development and application of mobile gas flowrate calibration equipment[J].Oil & Gas Storage and Transportation,2014,33(11):418.[doi:10.6047/j.issn.1000-8241.2014.04.016]
[4]廖柯熹,肖杰,朱勇,等.海底天然气管道多相流动与腐蚀状况分析[J].油气储运,2014,33(6):653.[doi:10.6047/j.issn.1000-8241.2014.06.018]
　LIAO Kexi,XIAO Jie,ZHU Yong,et al.Multiphase flow and corrosion analysis of subsea gas pipeline[J].Oil & Gas Storage and Transportation,2014,33(11):653.[doi:10.6047/j.issn.1000-8241.2014.06.018]
[5]冯亮.天然气管道投产充压时间的预测[J].油气储运,2016,35(1):59.[doi:10.6047/j.issn.1000-8241.2016.01.011]
　FENG Liang.Prediction of pressurizing time for commissioning of natural gas pipeline[J].Oil & Gas Storage and Transportation,2016,35(11):59.[doi:10.6047/j.issn.1000-8241.2016.01.011]
[6]郝晓晨.省域天然气产业发展思路探索[J].油气储运,2014,33(7):701.[doi:10.6047/j.issn.1000-8241.2014.07.004]
　HAO Xiaochen.Expected development of provincial gas industry[J].Oil & Gas Storage and Transportation,2014,33(11):701.[doi:10.6047/j.issn.1000-8241.2014.07.004]
[7]吴华,肖博文,彭宇,等.共用调压系统对特殊用户计量的影响[J].油气储运,2015,34(1):100.[doi:10.6047/j.issn.1000-8241.2015.01.021]
　WU Hua,XIAO Bowen,PENG Yu,et al.Effect of common pressure regulating system on metering for special users[J].Oil & Gas Storage and Transportation,2015,34(11):100.[doi:10.6047/j.issn.1000-8241.2015.01.021]
[8]王俊奇　李启锋　郑欣.天然气资源协调优化理论模型[J].油气储运,2016,35(预出版):1.
　WANG Junqi,LI Qifeng,ZHENG Xin.Natural gas resource coordination and optimization model[J].Oil & Gas Storage and Transportation,2016,35(11):1.
[9]狄向东张德文刘滨许庆元刘骁.天然气计量检定站施工工序的关键点[J].油气储运,2016,35(预出版):1.
　DI Xiangdong,ZHANG Dewen,LIU Bin,et al.Key points in the construction process of natural gas metrological verification stations[J].Oil & Gas Storage and Transportation,2016,35(11):1.
[10]刘金艳,侯磊,李丽敏,等.天然气地下储气库可靠性评价方法[J].油气储运,2015,34(2):171.[doi:10.6047/j.issn.1000-8241.2015.02.012]
　LIU Jinyan,HOU Lei,LI Limin,et al.Reliability evaluation method of underground gas storage[J].Oil & Gas Storage and Transportation,2015,34(11):171.[doi:10.6047/j.issn.1000-8241.2015.02.012]

更新日期/Last Update: 2023-07-25