面向泛在物联网的电力营销大数据处理方法 *

doi:10.11985/2020.01.014

摘要/Abstract

摘要：

针对目前电力营销管理数据存在互联性、实时性较差以及数据处理困难的问题,提出了新型的电力营销管理,将物联网技术、无线通信技术与大数据挖掘计算有效地结合起来,实现各个地区电力营销管理信息系统(Material management information system,MMIS)数据的互联、筛选与分析。采用码分多址(Code division multiple access,CDMA)、通用无线分组业务(General packet radio service,GPRS)、2G/3G/4G或无线调制解调器(Mobitex)等网络满足电力营销信息数据的远程无线通信,应用数据挖掘算法实现不同地区MMIS数据的筛选,在筛选时,通过建立决策树ID3算法模型,根据经验熵H(D)公式计算筛选属性,进而确定构建决策树的根节点和叶节点,从而构建出筛选多个MMIS数据的数学模型。又利用随机矩阵理论分析出不同地区MMIS数据之间的联系,有利于用户将各个地区MMIS数据内隐藏的宏观数据现象通过微观分析,进一步揭示出营销数据类型与各个地区MMIS之间的关联。试验表明,设计的方案大大改善了数据传输的效果,提高了数据管理能力。

关键词: 电力营销, 物联网技术, 无线通信技术, 数据挖掘算法, 随机矩阵理论

Abstract:

：Aiming at the problems of current power marketing management data, such as interconnectivity, poor real-time performance and difficult data processing, a new type of power marketing management is proposed, which effectively combines Internet of Things technology, wireless communication technology and big data mining calculation to realize various regions. Interconnection, screening and analysis of material management information system (MMIS) data. CDMA, GPRS, 2G/3G/4G or Mobitex networks are used to satisfy the remote wireless communication of power marketing information data. The data-mining algorithm is applied to realize the screening of MMIS data in different regions. During the screening, the decision tree ID3 algorithm model is established. The empirical entropy H(D) formula calculates the screening attributes, and then determines the root nodes and leaf nodes of the decision tree, thus constructing a mathematical model for screening multiple MMIS data. The random matrix theory is used to analyze the relationship between MMIS data in different regions, which is helpful for users to reveal the macro data phenomenon hidden in MMIS data of each region through micro analysis, and further reveal the relationship between marketing data types and MMIS in various regions. Experiments show that the scheme designed greatly improves the effect of data transmission and improves data management capabilities.

Key words: Power marketing, Internet of Things technology, wireless communication technology, data mining algorithm, random matrix theory

中图分类号:

TP311

彭龙,化振谦,党三磊,赵炳辉,张永旺. 面向泛在物联网的电力营销大数据处理方法 ^*[J]. 电气工程学报, 2020, 15(1): 95-102.

PENG Long,HUA Zhenqian,DANG Sanlei,ZHAO Binghui,ZHANG Yongwang. Power Marketing Big Data Processing Method for Ubiquitous Internet of Things[J]. Journal of Electrical Engineering, 2020, 15(1): 95-102.

图/表 14

图1

图2

图3

图4

图5

图6

图7

图8

表1

图9

表2

图10

图11

图12

参考文献 26

[1]	Meijer E . The world according to LINQ[J]. Communications of the ACM, 2011,54(10):45-51. doi: 10.1016/j.ahj.2018.10.002 pmid: 30487072
[2]	甄昊涵, 沈华 . 电能表现场检验数据挖掘研究[J]. 电子技术应用, 2017,43(4):76-78,82.
	Zhen Haohan, Shen Hua . Research on data mining of energy meter field inspection[J]. Application of Electronic Technique, 2017,43(4):76-78,82.
[3]	王远, 陶烨, 蒋英明 , 等. 智能电网时序大数据实时处理系统[J]. 计算机应用, 2015(S2):93-97.
	Wang Yuan, Tao Ye, Jiang Yingming , et al. Real-time processing system for smart grid sequential big data[J]. Journal of Computer Applications, 2015(S2):93-97.
[4]	蒋海明, 夏丹妮, 刘媛 , 等. 基于物联网技术的电力通信接入网的移动运维系统[J]. 电信科学, 2018,34(10):170-180.
	Jiang Haiming, Xia Danni, Liu Yuan , et al. Mobile operation and maintenance system for power communication access network based on Internet of Things technology[J]. Telecommunications Science, 2018,34(10):170-180.
[5]	杨凡, 王宜怀, 宋洪儒 . 基于窄带物联网应用架构的温湿度控制系统设计[J]. 电子技术应用, 2018,44(7):65-69.
	Yang Fan, Wang Yihuai, Song Hongru . Design of temperature and humidity control system based on narrowband IoT application architecture[J]. Application of Electronic Technique, 2018,44(7):65-69.
[6]	阿辽沙·叶 . 用电信息采集系统中多源异构通信协议设计[J]. 现代电子技术, 2018,41(2):145-147,151.
	Aliaosha Ye . Design of multi-source heterogeneous communi- cation protocol in electricity information acquisition system[J]. Modern Electronic Technology, 2018,41(2):145-147,151.
[7]	刘坤, 方芳, 王伟 . 基于MATLAB/RTW的通讯模块的设计与实现[J/OL]. 电子测量与仪器学报, 2015(2):296-301.
	Liu Kun, Fang Fang, Wang Wei . Design and implementation of communication module based on MATLAB/RTW[J/OL]. Journal of Electronic Measurement and Instrument, 2015(2):296-301.
[8]	张国强, 林永峰 . 智能电表检测系统通信协议栈设计与实现[J]. 现代电子技术, 2016,39(11):34-37,41.
	Zhang Guoqiang, Lin Yongfeng . Design and implement- tation of communication protocol stack for smart meter detection system[J]. Modern Electronic Technology, 2016,39(11):34-37,41.
[9]	王奇, 庄远灿, 阎帅 , 等. 基于随机矩阵理论的交直流输电通道线损大数据关联特性分析[J]. 电力自动化设备, 2018,38(5):77-83.
	Wang Qi, Zhuang Yuancan, Yan Shuai , et al. Correlation characteristics analysis of line loss big data of AC and DC transmission channels based on random matrix theory[J]. Electric Power Automation Equipment, 2018,38(5):77-83.
[10]	徐重酉, 韩翊, 贺兴 , 等. 基于随机矩阵理论的配电网阵列薄弱性评估系统设计[J]. 电器与能效管理技术, 2018(9):522-526,552.
	Xu Zhongyou, Han Yi, He Xing , et al. Design of array vulnerability assessment system for distribution network based on random matrix theory[J]. Electric Appliance and Energy Efficiency Management Technology, 2018(9):522-526,552.
[11]	陈星豪, 黄柳萍, 陶国飞 . 基于随机矩阵理论的电力大数据表征方法[J]. 电网技术, 2017,36(3):35-38.
	Chen Xinghao, Huang Liuping, Tao Guofei . A method for power big data characterization based on random matrix theory[J]. Power System Technology, 2017,36(3):35-38.
[12]	Li Q, Niu H, Papathanassiou A , et al. 5G network capacity:Key elements and technologies[J]. IEEE Vehicular Technology Magazine, 2014,9(1):71-78. doi: 10.1109/MVT.2013.2295070
[13]	涂莹, 朱炯, 裘华东 , 等. 面向“互联网+电力营销”的智能互动服务创新体系架构[J]. 中国电力, 2017,50(9):95-99.
	Tu Ying, Zhu Jiong, Qiu Huadong , et al. Intelligent interactive service innovation system architecture for “Internet+Electricity Marketing”[J]. China Electric Power, 2017,50(9):95-99.
[14]	刘素芳, 杨洋, 王溢泽 . 基于EPON技术的电力信息网架构分析与研究[J]. 电源技术, 2017,41(7):1068-1070.
	Liu Sufang, Yang Yang, Wang Yize . Analysis and research on the architecture of power information network based on EPON technology[J]. Power Technology, 2017,41(7):1068-1070.
[15]	国家电力公司发输电运营部. 电力营销管理信息系统设计规范(试行)[M]. 北京: 中国电力出版社, 2002.
[16]	范李平, 张晓辉, 苏伟 , 等. 基于大数据挖掘的变电设备故障预警研究及应用[J]. 电力大数据, 2019(1):7-13.
	Fan Liping, Zhang Xiaohui, Su Wei , et al. Research and application of fault early warning for transformer equipment based on big data mining[J]. Electric Power Big Data, 2019(1):7-13.
[17]	柳毅 . 基于数据挖掘的城市配电网故障定位方法[J]. 设备管理与维修, 2018(Z1):107-110.
	Liu Yi . Fault location method of urban distribution network based on data mining[J]. Plant Maintenance Engineering, 2018(Z1):107-110.
[18]	周云祥, 曹昉 . 基于数据挖掘的变电站监控后台告警信号自动分析[J]. 工业控制计算机, 2017,30(4):121-122,125.
	Zhou Yunxiang, Cao Fang . Automatic analysis of alarm signals in substation monitoring based on data mining[J]. Industrial Control Computer, 2017,30(4):121-122,125.
[19]	苏敬芳 . 高层建筑中低压配电系统的线路无功优化技术研究[J]. 科技通报, 2017,33(2):214-218.
	Su Jingfang . Research on line reactive power optimization techniques for low-voltage distribution system in high-rise buildings[J]. Bulletin of Science and Technology, 2017,33(2):214-218.
[20]	陈冰斌, 赵健 . 配电网无功补偿节能计算方法优化研究[J]. 电网与清洁能源, 2017,33(10):27-32,40.
	Chen Bingbin, Zhao Jian . Research on optimization of energy-saving calculation method for reactive power compensation in distribution network[J]. Power System and Clean Energy, 2017,33(10):27-32,40.
[21]	王磊, 陈青, 高洪雨 , 等. 基于大数据挖掘技术的智能变电站故障追踪架构[J]. 电力系统自动化, 2018,42(3):9-13.
	Wang Lei, Chen Qing, Gao Hongyu , et al. Intelligent substation fault tracing architecture based on big data mining technology[J]. Automation of Electric Power Systems, 2018,42(3):9-13.
[22]	韩晓慧, 杜松怀, 苏娟 , 等. 触电信号暂态特征提取及故障类型识别方法[J]. 电网技术, 2016,40(11):3591-3596.
	Han Xiaohui, Du Songhuai, Su Juan , et al. Method for extracting transient characteristics of electric shock signals and identifying fault types[J]. Power System Technology, 2016,40(11):3591-3596.
[23]	郝然, 艾芊, 肖斐 . 基于多元大数据平台的用电行为分析构架研究[J]. 电力自动化设备, 2017,37(8):26-33.
	Hao Ran, Ai Qian, Xiao Fei . Research on the analysis framework of electricity behavior based on multivariate big data platform[J]. Electric Power Automation Equipment, 2017,37(8):26-33.
[24]	贺红燕 . 基于大数据的智能电网关键技术研究[J]. 电源技术, 2016,40(8):1713-1714.
	He Hongyan . Research on key technologies of smart grid based on big data[J]. Power Technology, 2016,40(8):1713-1714.
[25]	刘鑫, 贾云献, 孙磊 , 等. 基于BP神经网络的变速箱故障诊断方法研究[J]. 计算机测量与控制, 2017,25(1):17-20.
	Liu Xin, Jia Yunxian, Sun Lei , et al. Research on gearbox fault diagnosis method based on BP neural network[J]. Computer Measurement and Control, 2017,25(1):17-20.
[26]	Slavakis K, Kim S, Mateos G , et al. Stochastic approximation vis-a-vis online learning for big data analytics[J]. IEEE Signal Processing Magazine, 2014,6(6):124-129.

数据库集合	数据类型	Class=1	Class=2	计算结果
D			8 7	H(D)=0.996 8
A₁	D₁	8	2 6	g(D, A₁)=0.288 0
A₁	D₂	7	6 1	g(D, A₁)=0.288 0
A₂	D₁	5	4 1	g(D, A₂)=0.139 8
	D₂	4	1 3
	D₃	6	3 3
A₃	D₁	3	1 2	g(D, A₃)=0.029 2
A₃	D₂	12	7 5	g(D, A₃)=0.029 2
A₄	D₁	7	6 1	g(D, A₄)=0.288 0
A₄	D₂	8	2 6	g(D, A₄)=0.288 0
A₅	D₁	6	5 1	g(D, A₅)=0.413 1
	D₂	4	0 4
	D₃	3	3 2

参数	采样周期/h	采样时间/h	D₁规模	D₂规模	D_std 计算结果
电费	8	0.5	80×150	400×500	0~45
收费	8	0.5	80×150	400×500	0~45
负荷	8	0.5	80×150	400×500	0~45