本文選題：多維度 + 聚類分析��； 參考：《湖南大學》2016年碩士論文

【摘要】：隨著智能電網(wǎng)的建設(shè),以智能電表為基礎(chǔ)的高級測量體系(Advanced Measurement Infrastructure, AMI)能夠延伸到普通用戶,智能電表采集的海量實時用戶用電數(shù)據(jù)中隱藏著用戶的用電行為習慣,針對智能電網(wǎng)建設(shè)過程中需要準確掌握居民用電特性的要求,對這些數(shù)據(jù)進行挖掘,研究用戶類型,可以幫助電網(wǎng)了解用戶個性化、差異化服務(wù)需求,為未來電力需求響應(yīng)政策的制定提供數(shù)據(jù)支撐。本文以民用非生產(chǎn)性負荷為研究對象,以環(huán)保、節(jié)能、經(jīng)濟、安全用電為目標,從時間維度、類屬維度、影響維度分別對用戶的用電特性進行分析,為需求響應(yīng)提供有效的數(shù)據(jù)支撐。本文的主要內(nèi)容如下：(1)對負荷特性指標進行了分類,并詳細介紹了各負荷特性指標的計算方法及其意義,為下文負荷特性的計算和負荷特性曲線的繪制奠定了基礎(chǔ)。從電價政策、電力供應(yīng)能力、需求側(cè)管理措施、氣溫氣候等方面進行了影響地區(qū)負荷的主要因素分析,同時介紹了負荷特性分析中常用的分析方法,并通過實例表明該方法在負荷特性精細化分析中存在不足。(2)將電力用戶負荷分解為基本負荷和季節(jié)性負荷,在利用自適應(yīng)模糊c均值算法對電力用戶基本負荷和夏季降溫負荷分別進行分類的基礎(chǔ)上,綜合考慮日用電量和電力負荷的峰谷特征,基于加權(quán)重心典型用戶篩選模型完成各類典型用戶的篩選,并利用灰色關(guān)聯(lián)度分析法對篩選的結(jié)果進行分析比較,結(jié)果表明該方法篩選的典型用戶是可行的、合理的、有效的。(3)計算各類典型用戶的日負荷率、峰期耗電率、平期耗電率、谷期耗電率等負荷特性指標,結(jié)合各類典型用戶基本負荷曲線和夏季降溫負荷曲線,分別分析基本負荷特性和夏季降溫負荷特性,并對其進行負荷調(diào)控潛力分析,提出了一種新的用戶分類方法。在電力用戶負荷依據(jù)調(diào)控潛力的大小重新分類的基礎(chǔ)上,從用電類型、負荷特性、錯峰潛力3個維度,分析各個用戶的用電行為特性,針對不同的用戶用電行為特性制定不同的錯峰管理政策,輔助需求側(cè)管理的實施。
[Abstract]:With the construction of smart grid, Advanced Measurement Infrastructure (Amis), an advanced measurement system based on smart meter, can be extended to ordinary users. In view of the requirement of accurately mastering the characteristics of household electricity consumption in the process of smart grid construction, mining these data and studying the types of users can help the power grid to understand the personalized and differentiated service requirements of users. To provide data support for future power demand response policy formulation. This paper takes the civilian non-productive load as the research object, taking the environmental protection, energy saving, economy, safety electricity consumption as the goal, from the time dimension, the category dimension, the influence dimension respectively carries on the analysis to the user's electricity use characteristic. Provide effective data support for demand response. The main contents of this paper are as follows: (1) the load characteristic index is classified, and the calculation method and significance of each load characteristic index are introduced in detail, which lays a foundation for the calculation of load characteristic and the drawing of load characteristic curve below. The main factors affecting regional load are analyzed from the aspects of electricity price policy, power supply capacity, demand-side management measures, temperature and climate, etc. At the same time, the analysis methods commonly used in load characteristic analysis are introduced. An example is given to show that the method has shortcomings in the refined analysis of load characteristics. (2) the power user load is decomposed into basic load and seasonal load. Based on the classification of basic load and cooling load in summer by using adaptive fuzzy c-means algorithm, the peak and valley characteristics of daily power consumption and power load are considered synthetically. Based on the typical user screening model of weighted gravity center, the selection of typical users is completed, and the results of screening are analyzed and compared by using grey correlation analysis. The results show that the typical users screened by this method are feasible and reasonable. The daily load rate, peak power consumption rate, average power consumption rate, valley power consumption rate and other load characteristic indexes are calculated effectively. Combined with the basic load curve and summer cooling load curve of various typical users, the basic load curve and the summer cooling load curve are used to calculate the daily load rate, the peak power consumption rate, the average power consumption rate and the valley period power consumption rate. The basic load characteristics and summer cooling load characteristics are analyzed, and the potential of load regulation is analyzed, and a new user classification method is proposed. Based on the classification of power user load according to the potential of regulation and control, this paper analyzes the characteristics of electricity consumption behavior of each user from three dimensions: power type, load characteristic and potential of wrong peak. According to different characteristics of consumer's power consumption behavior, different management policies are made to assist the implementation of demand side management (DSM).
【學位授予單位】：湖南大學
【學位級別】：碩士
【學位授予年份】：2016
【分類號】：TM714
，

本文編號：1857678