基于驱动因素分解的能源消费预测——以上海市为例  

Energy Consumption Forecast Based on Driving Factor Decomposition——a Case Study of Shanghai

作　　者：高迪 任庚坡[2] 李琦芬 毛俊鹏 桂雄威 GAO Di;REN Gengpo;LI Qifen;MAO Junpeng;GUI Xiongwei(Shanghai University of Electric Power, Shanghai 200090, China;Shanghai Energy Conservation Monitoring Center, Shanghai 200083, China)

机构地区：[1]上海电力大学,上海200090 [2]上海市节能监察中心,上海200083

出　　处：《重庆理工大学学报（自然科学）》2021年第9期269-277,共9页Journal of Chongqing University of Technology：Natural Science

基　　金：国家自然科学基金项目(71403162)。

摘　　要：将上海市工业能源消费增长的驱动因素分解为3部分,综合预测上海市工业能源消费总量需求趋势。采用ARIMA-BP神经网络组合模型预测工业企业能源消费量E1;情景分析法预测工业固定资产投资项目新增能源消费E2;历年统计数据预测产业结构调整及节能技改减少量E3。预测结果显示:组合模型能有效降低平均相对误差,提高预测精度,指标分析全面,适用于上海市工业能源消费预测。预测趋势表明:“十三五”后期上海市工业能源消费将呈现增长趋势,但增速较为平缓;进入“十四五”时期,上海工业能源消费总量将进一步放缓。This paper analyzes the driving factors of Shanghai’s industrial energy consumption growth in three parts and predicts the trend of Shanghai’s industrial energy consumption.The combined model of ARIMA-BP neural network was used to predict the energy consumption of industrial enterprises(E1).The scenario analysis method predicts new energy consumption of the industrial fixed assets investment projects(E2).The statistical data over the years predict the industrial structure adjustment and the reduction of energy-saving technology(E3).The prediction results show that the combined model can effectively reduce the average relative error and improve the prediction accuracy,and the index analysis is comprehensive.The forecast shows that the industrial energy consumption in Shanghai will increase in the later period of the 13th five-year plan,but the growth rate is relatively slow.Entering the period of the 14th five-year plan,Shanghai’s total industrial energy consumption will further slow down.

关 键 词：驱动因素分解 ARIMA-BP组合模型 能源消费预测 固定资产投资项目 

分 类 号：F201[经济管理—国民经济]