This paper proposes a new application of principal component regression (PCR) for estimating electrical energy consumption in case of abnormal automatic meter reading (AMR) systems. These events occur in a delivery metering system such as problems from mistakenly setting and connecting meters in electrical systems, broken metering accessories, etc. The estimation is performed by using MATLAB. The unclean sampled input data is used to estimate the target output data. The mean absolute percentage error (MAPE) is used as estimation performance. In this proposed estimation, load profiles obtained from the AMR are used as input data for training to create estimation model and for testing to validate model. Estimated results are verified by comparison between the proposed PCR application and other applications such as simple linear regression (SLR), multiple linear regression (MLR). The proposed PCR gives the best error results of MAPE for the lost electrical energy estimation.
Key words: Automatic meter reading (AMR), load profiles, principal component regression (PCR), multiple linear regression (MLR), simple linear regression (SLR).
Copyright © 2020 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0