Abstract: Electricity Theft has grown in tandem with the rise in electricity use. Electricity theft is causing widespread power outages and developing nations such as India are facing a major electricity crisis. When a consumer tampers with the units consumed it is called Electricity Theft. Because of illegal intervention with electric meters, power utilities are in deep financial trouble. In this paper, we are going to present a Desktop Application to Detect Electricity theft in Smart Distribution grids using machine learning. In this system, we utilize OCR (Optical Character Recognition) to recognize the meter reading (units used) from the meter image and convert it into machine-readable text. The SARIMAX (Seasonal Auto-Regressive Integrated Moving Average with eXogenous components) algorithm is a new variant of the ARIMA algorithm that is used to anticipate consumer consumption and detect Electricity Theft. If theft is detected, a message will be delivered to the fraudulent customer. The suggested method would assist Electricity Boards in detecting and recovering losses from electricity theft.
Keywords: OCR, SARIMAX, Electricity Theft, Electricity Board, Desktop Application, Machine Learning.
|
DOI:
10.17148/IJARCCE.2024.13582
