Power System Faults Detection Using Machine Learning with PMU Based Wide Area Monitoring

An effective control system is required to increase the power system’s safety, effectiveness, and reliability. Phasor measurement unit (PMU) and the wide area measurement systems (WAMS) both synchronized with the global position system can provide timestamp and synchronized phasors which provides a dynamic view of power system. Combining these synchronized measurements in a central protection system (CPS), a wide area monitoring system is created by means of optical fiber communication. Since PMU devices are more efficient than standard SCADA (Supervisory Control and Data Acquisition) systems at capturing the rapid dynamics of the power system, this technology is becoming popular in the utility industry. The suggested concept is designed to operate promptly for protecting large power transmission networks using PMUs. The proposed technique has been implemented on IEEE 5-bus network with MATLAB 2021a and has identified faults within a few milliseconds. The machine learning technique was used to identify anomalies, aiming to enhance the reliability of the power system. To analyze the massive amounts of PMU data produced in the power systems, this study has provided classification methods for K Nearest Neighbors, Logistic Regression method and Support Vector Classifier. These algorithms’ results indicate that they are quite accurate at identifying anomalies in the data provided by PMUs.