基于电力大数据的企业复工电力指数研究与应用Research and Application of Resumption Power Index Based on Power Big Data
尹积军,潘巍巍
YIN Jijun,PAN Weiwei
摘要(Abstract):
疫情期间,企业复工情况不仅关系到地方税收和就业,还能深刻反映地方的经济和社会运转状况。准确掌握地区企业复工情况,既可帮助地方政府有效推动复工复产,也可辅助政府制定针对性防疫措施。通过收集浙江省内企业近3年春节期间电量相关基础数据,提出了一种基于GMM(高斯混合模型)和Knee Point算法的企业复工电力指数计算方法。首先,利用GMM对企业进行聚类分析,将其分为春节停工企业和春节不停工企业;其次,利用Knee Point算法分析企业的用电曲线,对春节停工企业是否复工进行判断;最后,基于研究得到的企业复工电力指数模型计算浙江省内各地区复工电力指数,并将计算结果用于辅助政府进行复工复产决策。
Enterprise resumption of work amidst COVID-19 is not only related to a local tax revenue and employment but deeply reflects its economic and social operation status. Good mastery of regional enterprise resumption of work can help the local government promote the resumption of work and production and formulate targeted epidemic prevention measures. After basic data collection about electricity consumption of enterprises in Zhejiang province during the Spring Festival in the last three years, this paper proposes a calculation model for the resumption power index based on GMM(Gaussian mixture model) and Knee Point algorithm. Firstly, GMM is used to perform cluster analysis on the companies and divide them into two groups, namely one stops work and the other still in operation during Spring Festival. Secondly, the Knee Point algorithm is used to analyze the enterprise power consumption curve to judge whether the company resumes work. Finally, the resumption power index in various regions based on resumption power index model is calculated, and the calculation result is used to assist government decision-making for resumption of work and production.
关键词(KeyWords):
复工复产;电力大数据;企业复工电力指数
resumption of work and production;power big data;resumption power index
基金项目(Foundation): 国网营销专项基金(6211XT20002J)
作者(Author):
尹积军,潘巍巍
YIN Jijun,PAN Weiwei
DOI: 10.19585/j.zjdl.202102005
参考文献(References):
- [1]王晶.基于循环经济的企业运行机制、模式及评价研究[D].武汉:华中科技大学,2007.
- [2]李赋欣,罗晓伊,沈军.基于电力数据的经济景气指数模型研究[J].四川电力技术,2018,41(4):64-68.
- [3]谈一鸣,孙伟卿.基于经济分析的上海电力景气研究[J].电网与清洁能源,2016,32(7):45-50.
- [4]邓雪晴.基于电力数据的经济发展趋势分析[J].中国市场,2017(32):28-29.
- [5]田传波,尹玉,金鹏,等.基于电力数据影响城市宏观经济发展分析与预测[J].电气应用,2015,34(增刊1):232-233.
- [6]杨东伟.从电力数据看经济发展趋势[J].华东电力,2013,41(6):1292-1295.
- [7]李海.电力消费量与国房景气指数关系研究[J].统计与决策,2018,34(5):95-98.
- [8]刘玉娇,宋坤煌,王向.基于电力大数据的经济景气指数分析[J].电信科学,2020,36(6):166-171.
- [9]VELMURUGAN T.Efficiency of K-Means and K-Medoids algorithms for clustering arbitrary data points[J].International Journal of Computer Technology&Applications,2012,3(5):1758-1764.
- [10]张少敏,赵硕,王保义.基于云计算和量子粒子群算法的电力负荷曲线聚类算法研究[J].电力系统保护与控制,2014,42(21):93-98.
- [11]范霄文.基于粗糙集的定性数据分析方法研究[D].厦门:厦门大学,2008.
- [12]孙吉贵,刘杰,赵连宇.聚类算法研究[J].软件学报,2008,19(1):48-61.
- [13]郝占刚,王正欧.基于遗传算法和K-medoids算法的聚类新算法[J].现代图书情报技术,2006(5):44-46.
- [14]吴倩倩,何友全.基于K-medoids算法的RFAT客户细分[J].华北水利水电大学学报(社会科学版),2016,32(3):44-46.
- [15]姚丽娟,罗可,孟颖.一种新的K-medoids聚类算法[J].计算机工程与应用,2013,49(19):153-157.
- [16]曾利军,陈敏,罗细飞.改进蚁群化学聚类算法在短期负荷预测中的应用[J].电力系统保护与控制,2012,40(4):59-62.
- [17]张宜浩,刘智,朱常鹏.融合距离度量和高斯混合模型的中文词义归纳模型[J].计算机科学,2017,44(8):265-269.
- [18]资和周.优先聚类和高斯混合模型树相融合的递增聚类研究[J].现代电子技术,2017,40(19):177-181.
- [19]刘松吟.基于聚类分析的电力通信网络流量分类算法研究[D].南京:东南大学,2016.
- [20]尹楠.基于高斯混合模型的期望最大化聚类算法[J].统计与决策,2017(4):87-89.
- [21]曹振丽,孙瑞志,李勐.一种基于高斯混合模型的不确定数据流聚类方法[J].计算机研究与发展,2014,51(增刊2):102-109.
- [22]刘明辉,李炜.基于knee points的改进多目标人工蜂群算法[J].计算机工程与应用,2018,54(2):40-47.
- [23]FASHOTO S G,OLUMIDE O,JACOB G.Application of Data Mining Technique for Fraud Detection in Health In surance Scheme Using Knee-Point K-Means Algorithm[J].Australian Journal of Basic&Applied Sciences,2013,6(8):558-562.
- [24]MALTESE J,OMBUKI-BERMAN B M,ENGELBRECHT A P.Pareto-based many-objective optimization using knee points[C]//2016 IEEE Congress on Evolutionary Computation(CEC).Vancouver,BC,Canada:IEEE,2016:3678-3686.