中文摘要：

本文对各种研究方法进行分类比较，并指出它们之间的相通性，认为在实证研究中普遍存在由于经济部门、产业结构、能源种类等划分的宽口径带来的实验误差。宽口径处理方法的盛行一方面是由于详尽数据的获得存在难度；另一方面是因为国内外现有研究都是采取“单项连乘和”的形式，这极大限制了分解表达式的涵盖范围和因子种类。为了克服实证研究误差，本文提出将“单项连乘和”扩展成“多项连乘和加总”形式，并对这种新的方程形式进行因素分解的计算和求解，最后将新方法运用到北京市2000～2011年的能源平衡表中，分解终端能源消费，获得了细致合理的实验结果。

英文摘要：

This paper analyses various show the similarities among them. Then we ches always have experimental error by the methods classification comparison and point out the present empirical resear- classificatory wide caliber in economic sector, industry structure and energy structure. The prevalence of wide caliber ap- proach partly because there is the difficulty to obtain detailed data, also because the existing studies are taken the ＂sum of single variables production＂ as decompo- sition function, which greatly limits the scope of the expression and factor catego- ries. In order to overcome the empirical research error, the paper expands ＂Sum of Single Variables Production＂ to ＂Aggregate the Sum of Polynomial Multiplication＂ form and then we calculate and solve this new form. Finally, the new method is ap- plied to the 2000～2011 energy balance of Beijing to decomposition the total final energy consumption and we obtain detailed and reasonable results.