Research On Spatial Correlation And Differentiation Of Regional Energy Efficiency

Since the reform and opening up,China's economy grow rapidly,the demand for energy is increasing.Although our country has adopted many policies to control the total energy production,while increasing imports from abroad,it is still hard to compensate for the energy gap.As the new energy has not been exploited at this stage,it is imminent to improve the efficiency of energy use.After reviewing and analyzing the previous studies on the correlation and differentiation of energy efficiency,firstly,the energy efficiency theory and most of its models are introduced briefly.Due to the lack of some indicators in Tibet,this paper selects 30 provinces and regions to analyze the data.In this paper,global auto-correlation test statistic Moran'I is used to explore the global correlation.It is found that theenergy efficiency has a significant positive correlation in space.After discovering the global auto-correlation of energy efficiency,we find that the local spatial aggregation is obvious by using the local correlation statistic LISA and Local Moran index I.Moran scatter plot can be found in most of the provinces in the third quadrant and the first quadrant,that is,low value area and low value area for temporary,high value area and high value area similar to that area has a positive spatial auto-correlation value.Spatial regression analysis was carried out by using spatial cross-sectional regression and spatial panel regression to study the regional spatial correlation after discovering global and local spatial correlation.The results show that the spatial correlation coefficient is also tested by the significance test,indicating strong spatial correlation.Compared with the ordinary OLS regression model,it is found that the spatial lag model and the spatial error model have better simulation results.Then,spatial correlation analysis is conducted from spatial and temporal perspective using spatial Durbin error model.It is found that not only the regional relevance of energy efficiency is obvious,but also the strong regional correlation between explanatory variables is found.Regional heterogeneity exists due to the heterogeneity of the elements between regions.In this paper,the GWR model is adopted to analyze and compare the four-year sections.Using of Criterion DIFF to determine the local regression significance test,it is found that the four factors are very obvious local effect.At the same time,the residual test also proves that the effect is better than that of ordinary OLS regression analysis.Combining the empirical results above,per capita GDP,and the proportion of secondary and tertiary industries have a major role in promoting per capita energy consumption.Per million patents per capita and industrial raw material purchase price index mainly have inhibitory effects on the per capita energy consumption.And the proportion of the secondary and tertiary industries has the greatest impact on the per capita energy consumption.At the same time,it is found that these four factors have an adverse effect on the energy consumption per capita in some points in time and provinces.Based on the conclusions above,this paper put forward relevant policy recommendations.The innovation of this paper is to find out the spatial correlation from the spatial correlation exploratory analysis,and then use the spatial lag model and spatial error model to analyze from several cross sections.The spatial Durbin panel model takes into account the spatial lag effects of the dependent variable and the explanatory variable,and the spatial Durbin panel is rarely found in the literature on energy efficiency.Finally,GWR model was used to analyze the regional differences from the three cross-sectional contrasts,and to compare with the global regression.