Frequency Prediction of Power Systems in FNET Based on State-Space Approach and Uncertain Basis Functions

In this paper, we discuss the modeling and prediction of power frequency. Power frequency is one of the most essential parameters in the monitoring, control, and protection of power systems and electric equipments because when a significant disturbance occurs in a power system, the frequency varies in time and space. It is critical to employ a dependable model in order to optimize the efficiency and reliability of power systems in the Frequency Monitoring Network (FNET), and thus, prevent frequency oscillation in power grid. This paper describes the use of a state-space model and basis functions to predict power frequency. In the state-space method, expectation maximization (EM) and prediction error minimization (PEM) algorithms are used to dynamically estimate the model’s parameters. In the basis functions method, we employ random basis functions to predict the frequency. The algorithms are easy to implement online, having both high precision and a short response time. Numerical results are presented to demonstrate that the proposed techniques are able to achieve good performance in frequency prediction.