中文摘要：

本文以中国29个省份为研究对象，使用经过LMDI改进后的面板数据模型从省级层面对中国的能源回弹效应进行测算，并进行面板聚类分析，发现中国能源回弹效应在不同时间段出现不同波动性。根据区域经济发展水平和资源禀赋影响，中国各地区能源回弹效应出现较为明显的集聚特征：一类是经济发展水平最高的能源消费地区，这些省份同时也是主要的能源输入地区或能源相对匮乏地区；二类是资源构成较为单一、能源消费量较少的能源输入地区；三类是经济发展水平居中的能源输入地区；四类是经济欠发达的能源主要输出地区。

英文摘要：

From the perspective of 29 Chinese provinces, this paper calculated the rebound effect using LMDI decomposition in panel data model and conducted cluster analysis. This paper found that Chinese energy rebound effect showed different volatility in different time periods. Chinese regional energy rebound effect also reflected cluster- characteristic apparently, according to the level of regional economic development and resource endownment. The first group is energy consuming area, which has the highest level of economic development and has been major energy-importing area or energy relatively scarce area. The second group contains energy-importing areas, which have single resource constitution and low energy consumption. The third group contains the energy-importing areas which have the middle level of economic development. The forth group contains the energy-exporting areas which have less developed economies.