Platform for detection and monitoring of the devices connected to the power grid

Abstract

This paper summarizes the creation of a platform for detection and monitoring devices connected to the power grid. The platform will provide monitoring of the power grid, for the purpose of detecting connected devices, with regard, to their potential threat to the operation of critical systems. The monitoring and detection will be performed using data provided by power-line communication modems connected to the monitored power grid. Data provided by power-line communication modems are in the form of the frequency spectrum of the noise. The the frequency spectrum of the noise will be statistically described, and this statistical description of the data will be used for training the machine learning model and subsequent detection of connected devices based on machine learning model prediction. All calculations of the statistical description of data, creation of machine learning model and subsequent detection and monitoring will be processed locally on a microcomputer using custom-developed code.