Classification of customer behaviour based on smart meter data

Sheikh Yusoff, Sheikh Amir Asyraaf (2021) Classification of customer behaviour based on smart meter data. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Classification of customer behaviour based on smart meter data.pdf - Submitted Version Download (365kB)
	Text (Full Text) Classification of customer behaviour based on smart meter data.pdf - Submitted Version Restricted to Repository staff only Download (1MB)

Abstract

Smart grids contribute to balancing output, delivery and usage by gathering information on the network. There were massive amount of data collected by the smart meter. Thus, the information data must be classify first into a few cluster to make sure the process of balancing between supply and demand more accurate and efficient. The main objective for this project are to simulate and perform the classification of customer behaviour based on smart meter data and to analyse the result of the classification on the customer behaviour. There are a few techniques that can be used to achieve the objective such as k-means, hierarchical and natural jenks natural break. After brief research and comparison, the method of K-means clustering techniques was used to classify the customer behaviour in this project. As a result, the customer was classified into domestic cluster, commercial cluster and industrial cluster by doing 4 case study which is clustering daily profiles, clustering daily mean working days profiles, daily mean weekend profiles and weekly mean profiles.

Actions (login required)

View Item