Abstract: Different influencing factors have different degrees of influence on natural gas demand. In order to fully understand the above differences, the relative degree of incidence is used to quantitatively describe the degree of correlation between natural gas demand and influencing factors. According to the principle of importance, the stepwise regression method is used to successively test the statistical significance of the influencing factors, deeply dig out the effective driving factors, and construct the “best” stepwise regression C-D production model to predict the future natural gas demand trend. The results indicate that: (1)GDP is the most important factor affecting China's natural gas demand, and energy consumption structure is also one of the important effective driving factors; (2)Based on the “optimal” stepwise regression C-D production function model which is constructed by important and effective driving factors,it achieves the prediction of inflection point fitting of historical data and has good prediction performance. And the prediction results can be used as an important reference to determine the future natural gas demand; (3)It is predicted that China's natural gas demand will increase steadily from 2020 to 2030, but the growth rate will slow down. By 2030, natural gas demand will reach approximately 566 billion cubic meters, and the growth rate will drop to 4%.

Key words: demand forecast, influencing factors, correlation analysis, stepwise regression analysis, C-D production function

摘要： 不同的影响因素对天然气需求的影响程度存在差异,为充分认识上述差异,利用相对关联度定量刻画天然气需求量与影响因素之间的关联程度。依据重要性优先原则,采用逐步回归法逐次地对影响因素的统计显著性进行检验,深度挖掘有效驱动因素,构建“最佳”逐步回归C-D生产函数模型预测未来天然气需求走势。结果表明:(1)GDP是影响中国天然气需求的最主要因素,能源消费结构也是重要的有效驱动因素之一;(2)基于重要有效驱动因素构建的“最佳”逐步回归C-D生产函数模型实现了历史数据拐点拟合预测,具有良好的预测性能,预测结果可作为确定未来天然气需求的重要参考依据;(3)预测2020~2030年中国天然气需求量呈稳步增长趋势,但增速有所减缓,到2030年天然气需求量达约5660亿立方米,天然气需求量增长率降至4%。

关键词: 需求预测, 影响因素, 关联分析, 逐步回归分析, C-D生产函数