Real-Time Low-Frequency Oscillations Monitoring and Coherency Determination in a Wind-Integrated Power System

With the increase of renewables in the generation mix, the problem of low-frequency oscillations (LFOs) further escalates in modern power systems. This paper proposes a method for LFO modes monitoring and coherency identification using Phasor Measurement Units (PMUs) and Artificial Neural Network (ANN) in real time for a high wind-penetrated power system. The data for ANN training and testing is obtained from PMUs that are placed optimally in the system for complete system observability. The synchronously sampled PMU data is dimensionally reduced using Principal Component Analysis before using it to train the ANN. The proposed approach is verified on the IEEE New England benchmark system. The results obtained validate the effectiveness of the proposed approach in predicting system damping and coherency status with very less computational burden under varying operating conditions.