Estimation of Time-varying Frequency and its Rate of Change in Low-inertia Power Systems

In this paper, a hierarchical estimation scheme is designed to track the frequency and its rate of change of non-stationary power signals. The frequency is retrieved by a kernel-based parameter estimator in the first step. Subsequently, the frequency estimates are injected into a kernel-based numerical differentiator to extract its changing rate. Thanks to the deployed Volterra integral operator and suitably designed kernel-functions, the proposed estimator can achieve very fast convergence speed without compromising the robustness against noise. Therefore, the real-time estimates are able to follow the time-varying frequency and its rate of change with satisfactory accuracy. The effectiveness and robustness of the proposed method are verified by numerical experiments considering typical practical scenarios under the disturbance of noise. The results of the proposed method are compared with a highly-concerned quadrature phase-locked-loop (QPLL) method.