Analysis of Residential Load Data and Its Applications for Smart Grids
- Publication Type:
- Thesis
- Issue Date:
- 2020
Open Access
Copyright Clearance Process
- Recently Added
- In Progress
- Open Access
This item is open access.
Smart grids are equipped with advanced smart metering infrastructures that enable the two-way communication between end-users and utilities and record the consumption data of electricity customers. The gathered smart meter data have opened up possibilities for analyzing the consumption behavior of customers and understanding the underlying factors affecting it. However, the dimensionality of recorded data necessitates the use of data analysis techniques to extract valuable information from the load profiles. In this regard, this thesis utilizes various methods to analyze the consumption and survey data of residential electricity customers.
In the first two chapters, the concepts of smart metering, residential customers’ load patterns, and clustering of load data are extensively discussed, and a comprehensive discussion of the extant literature is presented.
Chapter 3 presents a comparative study of five main clustering approaches including K-means, fuzzy c-means, hierarchical, self-organizing map, and Gaussian mixture models for load pattern segmentation. Various parameters of each of these methods are explained in detail and their performances are compared using six cluster validity indexes. The obtained results are analyzed to find out the characteristic load shapes among the load curves of customers and to identify the main consumption patterns.
The problems of data deluge and residential DR establishment are addressed in Chapter 4 using a combination of symbolic aggregate approximation (SAX), as a suitable dimensionality reduction technique, a clustering algorithm, and the entropy concept. The use of SAX can assist in the clustering of residential load patterns, which usually display high variability. Moreover, the results are utilized for ranking the customers based on their stability in usage patterns over time which is beneficial for different DR programs.
In Chapter 5, both the consumption data and survey data of residential electricity customers are used to find out the effects of the households’ socio-demographic attributes and building characteristics on load patterns.
In Chapter 6, a combination of clustering algorithms and optimization models are used to design TOU tariffs for electricity customers. The problem is modeled as a mixed-integer linear programming problem with the objective of maximizing the profits of an electricity retailer that participates in different market settlements. The stochastic programming technique is used to address the uncertainties in future load and price.
Finally, in the last chapter, future directions in the analysis of smart meter data and the clustering of load patterns are briefly reported. Furthermore, the proposals for future work are elaborated in this chapter.
Please use this identifier to cite or link to this item: