基于卫星灯光数据的我国省域电力消费模拟研究  

Simulation Study of Provincial Electricity Consumption Based on Satellite Lighting Data

作　　者：马忠玉[1] 肖宏伟[1] 

机构地区：[1]国家信息中心,北京100045

出　　处：《中国能源》2017年第1期6-10,14,共6页Energy of China

基　　金：国家重点研发计划项目(编号:2016YFA0602601);国家自然科学基金项目(编号:71573062)

摘　　要：本文利用美国国家海洋和大气管理局发布的全球卫星夜间灯光数据,在通过相互校准、同年度融合、年度间系统校准等处理得到我国省域夜间灯光数据的基础上,构建了我国省域卫星夜间灯光亮度DN值与人均电力消费量之间的时空地理加权回归模型,并对我国省域电力消费量进行了模拟。模拟结果显示,无论是全国整体还是各省域,2006—2013年年均用电量模拟值与实际值较为吻合,全国整体年均用电量的相对误差为0.2%,大部分省域的相对误差在1%以内;分年度分省域用电量来看,大部分省域的相对误差在5%以内。运用卫星灯光数据可以较为准确快速地对我国省域电力消费量进行估算和预测,为卫星大数据时代"透过区域看全国,透过电力看经济"进行电力经济监测分析奠定了基础。Based on the global nighttime lighting data released by the US National Oceanic and Atmospheric Administration, this paper gets China＇s provincial nighttime lighting data after applying inter-calibration, intra-annual composition, inter-annual series correction and so on, then builds the spatiotemporal geography weighted regression model between DN value of China＇s provincial nighttime lighting density and power consumption per capita, at last, simulates China＇s provincial electricity consumption. The simulation results show that, whether it is the whole country or the provinces, the average annual electricity consumption in 2006-2013 is in good agreement with the actual value. The relative error of the whole country is 0.2%, and the relative errors of most provinces are within 1%. From the view of each province＇s annual electricity consumption, the relative errors of most provinces are less than 5%. The satellite lighting data can be used to estimate and forecast the power consumption of China＇s provinces in a more accurate and speedy way, and builds the foundation for the economic analysis of the electric power economy just like ＂looking at the whole country through the region and seeing the economy through electricity＂ in the era of satellite large data.

关 键 词：电力消费 DMSP/OLS数据 灯光亮度DN值 数据校准 时空地理加权回归 

分 类 号：F426.61[经济管理—产业经济]