基于DEA和ANN的我国各地区工业能源效率及其影响因素分析  

Analysis on Industrial Energy Efficiency and its Influencing Factors of China's Provincial Regions based on DEA and ANN

作　　者：贺勇[1] 马爱文[1] 

机构地区：[1]广东工业大学管理学院,广东广州510520

出　　处：《数学的实践与认识》2016年第9期87-96,共10页Mathematics in Practice and Theory

基　　金：国家自然科学基金(71303061;71301030);教育部人文社科研究项目(11YJCZH057)

摘　　要：采用DEA方法测算了2005-2012年我国大陆30个地区工业能源利用效率的平均水平,通过相关分析确定能源效率的影响因素,并运用多层感知器神经网络模型对工业能源效率的影响因素进行重要性分析.结果显示:1)除少数省份能源得到充分有效利用外,其余省份都存在能源投入的冗余,特别是中西部地区的投入冗余量较多,其节能潜力较大;2)技术水平、研发强度、劳动生产率、国有化程度、市场开放程度对能源效率有显著影响,产业结构和工业企业平均规模无显著影响;3)按照对能源效率的影响程度排序,依次是市场开放程度、技术水平、研发强度、劳动生产率、国有化程度.First,DEA model is used to measure the average industrial energy efficiency of 30 provincial regions in China's Mainland during the year 2005 to 2012,then correlation analysis is adopted to predetermine the influencing factors of energy efficiency,and finally,multilayer perceptron neural network model is utilized to analyze the effect of the influencing factors on energy efficiency.The results show that：1） Except for a few provinces,the great majority of provinces have redundancy on energy input,especially in the central and western regions,of which have more energy saving potential;2） Technology level,RD intensity,labor productivity,the degree of nationalization and the openness of the market have significant impact on energy efficiency,but industrial structure and the average size of industrial enterprises have little effect on energy efficiency;3） According to the significance of influence factors on energy efficiency,the order is as following：the openness of the market,technology level,RD intensity,labor productivity,the degree of nationalization.

关 键 词：工业能源效率 影响因素 DEA 神经网络 

分 类 号：F424.1[经济管理—产业经济]