Hybrid energy storage power allocation based on moving average filtering and VMD

The original power of wind power plant must be stabilized by energy storage system to meet the national standard of grid connection. If the power generated by the energy storage device does not match its own physical characteristics, there will be serious consequences such as overcharge, over-discharge and even device damage. In this paper, the moving average filtering method combined with particle swarm optimization algorithm is adopted to determine the grid-connected target power, which can increase the grid-connected smoothness and reduce the total energy burden and maximum instantaneous power value of the energy storage unit. Power calming tasks were then assigned with Variational Mode Decomposition (VMD). In this paper, after setting the fitness function that can reflect the above indexes, the optimal combination of filtering window length, VMD decomposition numerical mode, penalty factor and boundary frequency is determined by particle swarm optimization algorithm. Combined with the prior knowledge of wind power grid connection standard and healthy charging and discharging cycle of energy storage device, the range of each parameter is limited to accelerate the convergence speed. Simulation results show that the smoothness of grid-connection can be improved as much as possible and the distribution of power leveling tasks between energy storage devices is reasonable. The charge and discharge burden of the energy storage device is very balanced and the capacity demand is low. It is proved that the proposed method is reasonable and effective.