This project presents a framework for continuous FDI threat detection and mitigation in Modern EPES. The repository contains scripts to evaluate the performance of FDI attack on sensor measurements in smart grids, with an ML-enabled detection module, enabling live mitigation of FDI attacks.
Contributors: Prof. Dimitrios Pezaros (dimitrios.pezaros@glasgow.ac.uk) Awais Aziz Shah (awaisaziz.shah@glasgow.ac.uk), Tahira Mahboob (tahira.mahboob@glasgow.ac.uk)
The repository is extended from initial work by Cara Bowes at: https://github.com/carabowes/DAME-FDIA (Cara Bowes) Follow the instruction at the repository to obtain initial results and evaluations.
##Updated investigations: Run scripts in directory 'Updated_evaluation_TM' running end-to-end evaluations, and 'updates_TM' to obtain plots ML evaluation metrics for attack scenarios: random, standard, stealth, for feature extraction methods: residual-based features, standard error, and state-based features, evaluation of attacks on all buses in IEEE 9 Bus system