【摘要】：隨著營配一體化協(xié)同工作機制的建立和營配數(shù)據(jù)的打通,用電客戶的用電數(shù)據(jù)能夠與客戶檔案、繳費等數(shù)據(jù)關聯(lián)起來�？蛻舻挠秒姅�(shù)據(jù)中隱含著客戶的用電行為特征,對這些用電數(shù)據(jù)進行深度挖掘并研究客戶類型,可以幫助電網(wǎng)了解客戶的個性化、差異化服務需求,從而使電網(wǎng)公司進一步拓展服務的深度和廣度,為未來的電力需求側響應政策的制定提供數(shù)據(jù)支撐。本論文首先是基于大數(shù)據(jù)技術通過深入分析梳理,確定了用電行為分析的外部數(shù)據(jù)和內(nèi)部數(shù)據(jù)的數(shù)據(jù)來源,分析了海量數(shù)據(jù)存儲技術和海量數(shù)據(jù)預處理技術,其次,研究了用電行為分析的關鍵技術,主要包括聚類算法、最優(yōu)聚類評價算法、日期匹配算法、曲線相似性度量算法等用電行為分析算法。再次,構建用電分析模型構建方案,詳細描述了用電分析模型的建模思路,研究了主網(wǎng)的用電特征和用電行為,然后結合主網(wǎng)的用電模式研究主網(wǎng)該模式下海量客戶的用電行為,從而利用模式匹配技術匹配主網(wǎng)用電模式和該模式下海量客戶的用電模式,建立歷史日削峰填谷匹配關系,并通過實證研究對用電分析模型進行驗證。最后,對用電行為分析管理軟件進行系統(tǒng)架構設計、系統(tǒng)功能設計、系統(tǒng)數(shù)據(jù)庫設計,并闡述系統(tǒng)的實現(xiàn)成果。用電行為分析管理軟件可以比較準確的對用電客戶的用電行為進行分析,有助于電力公司進一步指導用戶個性化智能用電,對提高電網(wǎng)需求側能效管理水平具有建設性意義。
[Abstract]:With the establishment of the cooperative working mechanism of operation and distribution integration and the opening of the distribution data, the electricity consumption data of the electric customers can be associated with the data of the customers' files, payment and so on. The customer's electricity consumption data is implicit in the customer's electricity behavior characteristic. It can help the power grid to understand the customer's individualized and differentiated service demand by digging these data deeply and studying the customer's type. So the grid company can further expand the depth and breadth of the service, and provide data support for the future DSM policy formulation. Firstly, based on big data technology, the external and internal data sources of electrical behavior analysis are determined, and the mass data storage technology and mass data preprocessing technology are analyzed. The key technologies of electrical behavior analysis are studied, including clustering algorithm, optimal clustering evaluation algorithm, date matching algorithm, curve similarity measurement algorithm and so on. Thirdly, the construction scheme of power consumption analysis model is built, the modeling idea of power consumption analysis model is described in detail, and the characteristics and behavior of power consumption of main network are studied. Then combining the power consumption mode of the main network, the paper studies the power consumption behavior of the massive customers under the main network mode, and then uses the pattern matching technology to match the power consumption mode of the main network and the massive customers under this mode, and establishes the matching relationship between the historical peak cutting and filling valley. And through the empirical study to verify the electricity analysis model. Finally, the system architecture design, system function design, system database design, and the achievement of the system are described. The analysis and management software of power consumption behavior can accurately analyze the power consumption behavior of customers, which is helpful for power companies to guide users to use electricity for personal intelligence, and to improve the management level of energy efficiency on the demand side of power grid.
【學位授予單位】：華北電力大學(北京)
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TP311.13;TM714;TM732

【參考文獻】

相關期刊論文 前10條

1 張蕾;崔志坤;李井泉;白濤;;改進K-means模型在電力系統(tǒng)用戶行為分析中的應用[J];無線電工程;2017年03期

2 王歡;李紅輝;張駿溫;;改進K-means聚類的云任務調度算法[J];計算機與現(xiàn)代化;2017年02期

3 陳鵬偉;陶順;肖湘寧;李璐;張劍;;短時間尺度用電行為相關性分析網(wǎng)絡模型[J];電力系統(tǒng)自動化;2017年03期

4 劉云霞;;基于動態(tài)時間規(guī)整的面板數(shù)據(jù)聚類方法研究及應用[J];統(tǒng)計研究;2016年11期

5 張素潔;趙懷慈;;最優(yōu)聚類個數(shù)和初始聚類中心點選取算法研究[J];計算機應用研究;2017年06期

6 莊緒強;;基于云計算技術的用戶用電智能分析技術研究[J];自動化與儀器儀表;2016年02期

7 戴月明;張朋;吳定會;;基于密度檢測的EM算法[J];計算機應用研究;2016年09期

8 伍育紅;;聚類算法綜述[J];計算機科學;2015年S1期

9 王繼業(yè);季知祥;史夢潔;黃復鵬;朱朝陽;張東霞;;智能配用電大數(shù)據(jù)需求分析與應用研究[J];中國電機工程學報;2015年08期

10 余冬梅;;基于障礙約束的空間聚類算法綜述[J];計算機系統(tǒng)應用;2015年01期

相關碩士學位論文 前3條

1 陳旭;基于神經(jīng)網(wǎng)絡的電力系統(tǒng)短期負荷預測研究[D];華中科技大學;2015年

2 潘家騰;基于Hive數(shù)據(jù)倉庫的用戶行為模型研究[D];北京郵電大學;2015年

3 冀明;聚類分析在電力客戶關系管理中的應用研究[D];華北電力大學;2012年

，

本文編號：2278025