EL-FAM: Power System Intrusion Detection with Ensemble Learning for False Alarm Mitigation

Power systems must be secured against malicious activity and unauthorized access with the help of intrusion detection systems (IDS). Unfortunately, it is often difficult for conventional IDS techniques to correctly identify robust and dynamic threats. The security of the power grid is seriously threatened by the growing frequency of attacks. In this paper an ensemble learning framework is proposed for power system intrusion detection in order to overcome this difficulty. This framework achieving high accuracy by combining the Random Forest, Decision Tree, and Logistic Regression algorithms. The proposed ensemble model is able to achieve an accuracy of 98%, which can be used to improve IDS outcomes [1]. Additionally, this method produces a low false alarm rate, that ensures precise and accurate intrusion detection. The system precisely detects intrusions by continuously observing real-time data for anomalies. Ensemble learning can incorporate data compression methods such as feature selection which reduce the overall memory footprint. This model's evaluation depicts that it has the potential to detect attacks on power systems, making it a useful tool for boosting overall security.