Research on the propagation characteristic of non-stationary wind power in microgrid network based on empirical mode decomposition

This paper researches on the inherent characteristic of the fluctuant wind power and its propagation characteristic in a microgrid simulation system. First, the wind power data are non-stationary random signals due to the wind's inherent characteristic, and are hard to describe by function. The non-stationary wind power data are decomposed into the trend part and stationary part by the empirical mode decomposition method. The stationary part is then decomposed into different instantaneous frequency components by Hilbert transform. Second, a low voltage microgrid network is built on the PSCAD/EMTDC, and it is also connected to a medium voltage distribution network. All kinds of components are inputted into the simulation system as the port property of the wind power generation system. The voltage fluctuation of some important buses are inspected and compared when each wind power component is inputted. Last, according to the impact of each component the fragile bus and the key component are found. The fragile bus is an important factor when locating the wind power system, and the key component of the wind power can guide the optimal control strategy which suppresses the fluctuation of wind power, for instance, only considering the key component can be more economic when using energy storage system to suppress the fluctuation.