Skip to Main Content
An improved Prony method for online dynamic tracking of low-frequency oscillations (DTLFO) in wide-are measurement system (WAMS) is proposed in this paper. It employs the characteristic of singular value distribution of the sample matrix to identify a reduced-order signal model. Then, the corresponding linear-prediction parameters are worked out in sense of singular linear least square. The obtained low-order system retains the dominant swing modes of interest and exhibits modal characteristics similar to the unreduced system in frequency domain associated with swing modes. Digital simulations verify the accuracy, speed and robustness of the improved algorithm. In one of the latest commissioned WAMS project, i.e., Jiangsu provincial WAMS, this algorithm is implemented as an online application to dynamically track low-frequency power oscillation in real time. The captured oscillatory event demonstrates the feasibility and good performance of the algorithm. With the reduced-order system model identified using the improved Prony method, our next work is to develop an online adaptive damping control system based on wide-area measurements.