Correlation between predictability index and the error performance of customer baseline load (CBL) calculation
- Published
- Accepted
- Subject Areas
- Adaptive and Self-Organizing Systems, Data Mining and Machine Learning, Data Science
- Keywords
- Randomized Controlled Trial (RCT), Discrete Fourier Transform (DFT), K-means clustering, CAISO, Predictability Index
- Copyright
- © 2016 Mohajeryami et al.
- Licence
- This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ Preprints) and either DOI or URL of the article must be cited.
- Cite this article
- 2016. Correlation between predictability index and the error performance of customer baseline load (CBL) calculation. PeerJ Preprints 4:e2374v1 https://doi.org/10.7287/peerj.preprints.2374v1
Abstract
This paper attempts to explore the correlation between the content of high frequency component of customers' historical consumption data (measured by a proposed index called predictability index) and the accuracy of Customer Baseline Load (CBL) calculation methods. In this paper, the customer's consumption signal is transformed from time-domain to frequency domain to separate the high and low frequency components of the consumption signal. Then, after reconstructing the time-domain equivalent of both of these signals, the predictability index for all customers are calculated. The data employed by this study belong to Australian Energy Market Operation (AEMO), and is the hourly consumption of 189 customers for the time span of a year (2012). This index is proposed to be used for the purpose of clustering the customers into different bins by K-means clustering algorithm. Then the CBL for customers of each bin is calculated by two methods of CAISO and Randomized Controlled Trial (RCT), and then the average error in each bin is computed. Afterwards, the correlation between the average P_index of each bin, and its normalized average error is calculated. It is found that there is a strong correlation between the P_index and the error performance of the CBL calculation methods.
Author Comment
This is a preprint submission to PeerJ.