区域能源效率的空间关联及分异研究

发布时间：2018-04-14 14:23

本文选题：能源效率 + 空间相关性　；参考：《浙江工商大学》2017年硕士论文

【摘要】：自改革开放以来,随着经济总量与实力的不断增长,经济增长对能源的需求不断增加,石油与煤炭资源过渡使用。虽然国家近些年来出台许多政策控制能源生产总量,同时加大国外进口,但是仍旧难以弥补能源的缺口。在新能源未被发现利用之前,提高能源利用效率迫在眉睫。在回顾及分析已有学者对能源效率的关联及分异性研究后,在此基础上,对能源效率的理论及其模型进行阐释。本文选择30个省份地区(除西藏)的数据进行分析。首先采用全局自相关检验统计量Moran'I值进行全局关联探索,发现能源效率在空间上有十分显著的空间正相关性。然后再利用局部空间相关统计量LISA值及其显著性检验,发现局部空间聚集现象明显。通过Moran散点图的结果,可以发现大部分省份位于第三象限与第一象限,即低值省份与低值省份为临,高值省份与高值省份相近,说明地区数值存在正空间自相关性。在发现存在明显的全局及局部空间关联性后,利用空间截面回归模型及空间杜宾误差模型对研究地域进行空间计量回归分析。结果显示空间影响系数通过显著性检验,说明空间关联性较强。与普通OLS回归模型比较后发现,空间滞后模型与空间误差模型的模拟效果更佳。接着利用空间杜宾误差模型从时空角度分析其空间关联性,发现不但能源效率的地区关联性明显,同时发现解释变量及误差项之间也有较强的地区关联性。由于地区之间的要素不匀质性,导致地区差异性存在。本文采用GWR模型对四年的截面进行分析对比。利用DIFF of Criterion对局部回归系数的显著性进行检验判断,发现大部分影响因素都很明显,说明局部回归效果优于全局回归,同时残差检验也证明GWR模型效果优于普通OLS回归分析。结合以上实证结果发现,人均GDP,人均专利数及工业生产者购进价格指数的增加对能源使用效率的提高主要具有正向促进作用,但空间杜宾误差模型中工业生产者购进价格指数的溢出效应系数为正值,说明周围地区价格的上升对本地的能源效率有抑制作用。第二、三产业比重的增加对本地的能源使用效率的提高具有抑制作用,但在空间杜宾误差模型中其溢出效应系数为负值,即周围地区该要素的增加促进本地能源使用效率的提高。在GWR模型中,在2008年受全球金融危机影响,所有地区工业生产者购进价格指数的系数均为正值,即对能源使用效率的提高具有抑制作用。本文同时发现各地区其他影响能源效率因素的禀赋不同,在空间上也具有明显溢出作用。基于上述结论提出了相关政策建议。本文主要创新点:利用空间杜宾误差模型同时将选取的四个指标的空间滞后影响考虑在内,同时探讨能源要素的直接与溢出效应。然而这些要素并不能完全反应影响能源使用效率的具体情况,各地区由于其他影响能源效率因素的禀赋不同,也会对周围地区有溢出影响,空间杜宾误差模型可以将这些因素放在误差项里进行整体分析。
[Abstract]:Since the reform and opening up, with the continuous growth of the total economy and the strength of the economic growth and the increasing demand for energy, petroleum and coal resources in transition countries in recent years. Although the introduction of many policies to control the total energy production, while increasing foreign imports, but still difficult to compensate for the energy gap. Before the new energy has not been found by and improve the efficiency of energy use is imminent. In the review and analysis of scholars related to energy efficiency and specific research, based on the interpretation of the theory and model of energy efficiency. This paper chooses 30 provinces (except Tibet) data were analyzed. Firstly, using global autocorrelation test statistic Moran'I value the global association of exploration, found that energy efficiency has a very significant positive correlation between space and space. And then using local spatial correlation statistics and LISA value The test of significance, find local spatial aggregation phenomenon is obvious. The Moran plot results can be found in most of the provinces in the first quadrant and third quadrant, namely low value and low value for the provinces of provinces, provinces of high value and high value provinces are similar, there is a numerical description of regional spatial autocorrelation. In found obvious the global and local spatial correlation, regression model and spatial Durbin error model for the study of regional spatial econometric regression analysis using the space section. The results showed that the coefficient of spatial effects through the significant test, the spatial relevance of the discovery and the ordinary OLS regression model. After comparison, better simulation results and the spatial error model of spatial lag model the use of space. Then the Durbin error model to analyze the spatial relationship from the angle of time and space, we found that not only the area association of energy efficiency was also found to explain There are also regional strong correlation between variables and the error term. Because of elements between regions is not uniform, leading to the existence of regional differences. This paper uses a comparative analysis on the cross section of four years of the GWR model. Of Criterion was on the local regression coefficients were tested by judge DIFF, found that most influence factors are obviously, that is better than the global local regression regression, and residual test also proved that GWR model is better than the effect of general OLS regression analysis. According to the above empirical results, the per capita GDP, per capita number of patents and industrial producer price index increased mainly has positive effect on energy use efficiency, but the spillover coefficient spatial Durbin error the model of industrial producer price index is positive, indicating the area around the price rise has inhibitory effect on local energy efficiency second, third. Industry's increasing use efficiency of local energy increase with inhibition, but in space Durbin error model and its spillover effect coefficient is negative, increasing the elements of the surrounding area to promote local energy use efficiency. In GWR model, in 2008 by the global financial crisis, the purchase price index area of industrial producers coefficients were positive, which has inhibitory effect on energy use efficiency. This paper also found that the influence of other area energy efficiency factor endowment different, also has obvious spillover effects in space. Based on the above conclusions relevant policy recommendations were put forward. The main innovation of this paper: four indicators using spatial Durbin at the same time the error model of the spatial lag effect into account, and to explore the direct and spillover effects of energy factors. However these factors can not completely reverse Effect of specific energy efficiency, the energy efficiency effect due to other factors endowment different, will also have a spillover effect on the surrounding area, spatial Durbin error model of these factors can be put in error in the overall analysis.

【学位授予单位】：浙江工商大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F426.2

【参考文献】