Abstract: Based on the panel data of30 provinces in China from2004 to2016,this study used SPSS software to conduct K-means cluster analysis on urbanization level and energy consumption structure,and empirically test the mechanism of urbanization level on energy consumption structure by using spatial Durbin model.The results showed that,the level of population urbanization and land urbanization had the significant positive effect on the optimization of energy consumption structure;the energy endowment had the certain degree of inhibition on the transformation of energy consumption structure due to the dependence of energy endowment on energy.

Key words: energy consumption structure, urbanization level, K-means cluster analysis, spatial Dubin model