摘要
针对目前我国能源消费量预测中变量选取单一、预测误差较大等问题,结合我国能源消费实际情况,引入GDP、人口、煤炭消费量、通货膨胀率建立了我国能源消费量预测的多元回归模型。利用该模型对1985—2010年能源消费量进行拟合,拟合效果较好;利用该模型对2011—2013年能源消费量进行实际预测,最大误差为1.708%,平均误差为1.3269%,最小误差仅为0.6748%,预测精度较高,为我国能源消费量预测提供一种新的途径。
Aiming at the prediction of energy consumption in China in the variable selection of single, forecast errors, combined with the actual situation of Chinese energy consumption, the introduction of GDP, population, coal consumption, inflation was a multiple regression model to predict the amount of energy consumption in China. By using this model in 2010 energy consumption to 1985 ,the fitting effect was good. Using the model for forecasting energy consumption from 2011 to 2013, the maximum error was 1. 708% , the average error was 1. 3269%, the minimum error was only 0. 6748 % , high prediction accuracy. It provided a new way to China energy consumption prediction algorithms.
出处
《资源开发与市场》
CAS
CSSCI
2014年第11期1317-1319,共3页
Resource Development & Market
基金
辽宁省自然科学基金项目(编号:131138)
关键词
能源消费量
多元回归分析
时间序列
预测
energy consumption
multiple regression analysis
time series
prediction