Fast Frequency Derivative Estimator for Inverter-Based Distributed Energy Resources

The proliferation of inverter-based distributed energy resources (DER) makes the interfaced grid face a fast frequency response, and the resulted time-variant frequency features a considerable fluctuation that degrades the frequency stability. This frequency information can be estimated by the phase-locked loops/frequency-locked loops (PLLs/FLLs) embedded in the DER inverters, yet they are difficult to track this frequency and obtain its derivative quickly. This paper aims to tackle this problem. Firstly, a time-variant parameter model is proposed to reflect an accurate relationship between the frequency derivative and the input signal. Since the input signal is defined with ramp frequency, the given model overcomes the inherent time-invariant restriction of the well-known notch filter/model. Then, a nonlinear estimator is derived by the equality-constraint optimization to obtain the model's parameters. The nonlinear estimator avoids the slow convergence speed and biased time-variant parameter tracking induced by the linear estimation framework. The derived linearized model further proves that the proposed estimator can accurately track ramp frequency within 20 ms. Experimental tests under time-variant frequency disturbances verify the proposed estimator's faster and more accurate estimation performance than the commonly-used/recently published PLLs/FLLs.