Microgrid frequency regulation involving low-wind-speed wind turbine generators based on deep belief network

Abstract

With the development of low-wind-speed technology, it becomes a trend that low-wind-speed wind turbine generators (LWTGs) are integrated into a microgrid. However, the frequency stability of the microgrid has thereby been challenged since the increased penetration of wind power lowers the inertia of the microgrid. In order to investigate how LWTGs can effectively participate in suppressing the frequency fluctuation of the microgrid, virtual inertia control, over-speed control, as well as droop control, is applied to LWTG. Moreover, the de-loading ratio of over-speed control, along with the control parameters of virtual inertia control and droop control are all optimised under different wind speeds by virtue of the deep belief network, whereas the problem of over-speed control failure with the scheme of fixed de-loading ratio becomes more pronounced under low-wind speeds, which is defined as a blind area problem. To solve this problem, on the one hand, the strategy of the variable de-loading ratio is adopted under low-wind-speeds. On the other hand, the concepts of the minimum and maximum critical wind speed are deduced through theoretical analysis, which greatly restricts the number of feasible solutions of de-loading ratio under different wind speeds so as to improve the optimisation efficiency about 50%.