A Feedforward Neural Network Hydrogen Electrolyzer Output Regulator for Wind Power Control with Battery Storage

Wind power is a widely utilized renewable energy resource, but the irregular nature of wind affects the resultant power supply quality and results in fluctuating power. Large scale penetration of variable power causes issues such as frequency variations and voltage alterations. Battery Energy Storage Systems (BESS) are effectively used to smooth out the power variations and to resolve the fluctuations caused issues. However, a suitable control strategy is required to regulate the wind power output and battery state of charge (SoC) levels as it otherwise leads to the requirement of larger storage systems. This paper proposes a novel feedforward neural network controller (FNNC) for wind power and SoC control through the regulation of the hydrogen electrolyzer (HE) fuel cell (FC) system output based on the battery parameters feedback. The proposed FNNC takes the fluctuating wind power and battery parameters such as SoC and charging/discharging power as inputs and intelligently controls the HE output while complying with the imposed controller constraints. Simulation results conclude that the FNNC considerably lessens the wind power fluctuations and a comparison study between the widely used fuzzy logic control for SoC management demonstrates that the FNNC has better SoC reduction capabilities. The proposed controller also significantly reduces the ramp rate and promotes the incorporation of green hydrogen into the energy mix via the HE-FC system as a way for delivering controlled wind power into the grid with battery life enhancement through SoC management.