Detection of Data Integrity Attack in Cyber Physical Power System using Data-driven Method

A modernized and integrated electrical power grid called a cyber-physical power system (CPPS) combines conventional electrical infrastructure with cutting-edge digital communication, control, and monitoring technology. These systems use automation, data analytics, and the Internet of Things to improve the efficiency, reliability, and security of the production, distribution, and use of electricity. In order to guarantee system security and dependability, sensitive data must be protected from Data Integrity Attack (DIA). If the attack is not reported, a tragic occurrence will take place. For the operator in the control center to make an important decision on grid stability, DIA detection is therefore vital. This study discusses the impact of attacks on Wide Area Control (WAC) applications and data-driven methods of attack detection. The optimum detection technique is determined by comparing several detection approaches based on the validation of performance metrics. Based on simulation findings, it can be concluded that the Random Forest classifier performs better at detecting anomalies than other Machine Learning (ML) techniques like Support Vector Machine (SVM) and K Nearest Neighbour (KNN) classifier. The proposed technique is tested for effectiveness on the WSCC 3 machine 9 bus system.