Preventive frequency control of a microgrid with deloaded wind turbines using analytical frequency predictor and a PSO-based pitch angle optimizer

Microgrids that rely heavily on photovoltaic (PV) and wind generation are vulnerable to frequency dips under severe weather conditions. This paper presents a preventive frequency control strategy to avoid under-frequency load-shedding due to drastic changes in renewable power generation and load. An analytical frequency predictor based on series expansion of matrix exponential is first presented to estimate future system frequency accurately in an efficient manner. When there is risk that future system frequency will drop to a level below the under-frequency threshold, the pitch angle is increased in advance according to a trapezoidal pitch angle command such that mechanical energy can be stored in the de-loaded wind turbine. The reserved mechanical energy is released later on to improve future frequency nadir to a level above the under-frequency threshold. The parameters for the trapezoidal pitch angle command such as time and maximum pitch angle are determined using particle swarm optimization (PSO) algorithm. Results from MATLAB/Simulink simulations demonstrated the accuracy of the proposed analytical frequency predictor and the effectiveness of the proposed preventive frequency controller in avoiding future under-frequency load-shedding.