Network Submission and Monitoring of Power System Operation Safety Monitoring Data using Safe Power Hybrid Classifier (SPHC)

Abstract

Ensuring the dependability and security of power systems is critical to guaranteeing uninterrupted electricity supply and avoiding possibly catastrophic outages. Previous power system monitoring methods frequently have restricted incorporation and lack real-time assessment abilities, limiting their ability to detect and tackle safety problems promptly. To tackle these drawbacks, this paper presents the SafePower Hybrid Classifier (SPHC) algorithm, which is intended to improve real-time surveillance and classification of power system security information. The proposed system gathers and sends key security metrics like voltage, current, temperature, humidity, power factor, frequency, and phase imbalance from a network of sensors spread across power plants and transmission lines. The SPHC algorithm uses an ensemble voting method to classify the safety status as Normal, Warning, or Critical, enabling timely intervention using the Power System Operation Safety Monitoring Dataset (PSOSMD). Experimental findings indicate that the SPHC algorithm surpasses conventional classifiers, with performance metrics such as accuracy of 92.4%, precision of 91.3%, recall of 91.8%, F1-score of 91.5%, and MCC of 89.7%, substantially decreasing the possibility of power system failure. Integrating thorough data gathering and sophisticated classification into the proposed system greatly enhances the dependability and stability of power operations.