Smart Meter Data Analysis for Power Theft Detection

We propose a method for power theft detection based on predictive models for technical losses in electrical distribution networks estimated entirely from data collected by smart meters in smart grids. Although the data sampling rate of smart meters is not sufficiently high to detect power theft with complete certainty, detection is still possible in a statistical decision theory sense, based on statistical models estimated from collected data sets. Even without detailed knowledge of the exact topology of the distribution network, it is possible to estimate a statistical model of the technical losses that allows indirect estimation of the non-technical losses (power theft) with high accuracy. International Conference on Machine Learning and Data Mining in Pattern Recognition (MLDM) This work may not be copied or reproduced in whole or in part for any commercial purpose. Permission to copy in whole or in part without payment of fee is granted for nonprofit educational and research purposes provided that all such whole or partial copies include the following: a notice that such copying is by permission of Mitsubishi Electric Research Laboratories, Inc.; an acknowledgment of the authors and individual contributions to the work; and all applicable portions of the copyright notice. Copying, reproduction, or republishing for any other purpose shall require a license with payment of fee to Mitsubishi Electric Research Laboratories, Inc. All rights reserved. Copyright c ©Mitsubishi Electric Research Laboratories, Inc., 2013 201 Broadway, Cambridge, Massachusetts 02139