摘要
天然气作为一种优质、高效、清洁的低碳能源,近年来在我国能源结构中所占比重逐年加大,"十二五"规划明确将攻克天然气开发关键技术作为主要任务,以此为"十三五"页岩气大规模开发奠定基础。在此背景下,本文首先利用通径分析筛选出天然气消费的核心影响因素,发现人口和城镇化率是天然气消费的主要推动因素,GDP是天然气消费的主要限制因素。然后,运用智能算法RBF神经网络分位数回归(RBF-QRNN)模型对我国天然气消费进行分析和预测,结果表明"十二五"末,中国天然气消费量将近178 532.1百万立方米,2020年中国天然气消费量将近261 853.0百万立方米。
As a high quality,efficient,clean and low-carbon source of energy,share of natural gas in China's energy structure has been rising in recent years.The Twelfth Five-Year Plan has specified overcoming the critical technological challenges of natural gas exploitation as the main task,so as to lay the foundation for large-scale development of shale gas during the Thirteenth Five-Year Plan.Under this background,path analysis are used for screening the core of factors that influence natural gas consumption in the RBF neural network quantile regression(RBF-QRNN)model to predict gas consumption in China.The result shows the population and urbanization rate are push factors,while GDP is limiting factor.The natural gas consumption in China will reach around 178532.1million cubic meters by the end of'Twelfth Five-Year Plan'and about 261853.0million cubic meters in the year of 2020.
出处
《中国管理科学》
CSSCI
北大核心
2015年第S1期823-829,共7页
Chinese Journal of Management Science
基金
国家自然科学资助基金(71473155)
博士后基金特别资助项目(108633)
陕西师范大学中央高校特别资助项目(14SZTZ03)
关键词
天然气消费
通径分析
RBF神经网络
分位数回归
概率密度预测
natural gas consumption
path analysis
RBF neural network
quantile regression
probability density forecast
