An Intelligent Model Predictive Control Strategy for Stable Solar-Wind Renewable Power Dispatch Coupled with Hydrogen Electrolyzer and Battery Energy Storage

Renewable energy sources such as photovoltaic (PV) and wind power are widely used; however, their intermittent nature impairs power supply quality by creating frequency distortions and irregularities in voltage. Battery energy storage systems (BESS) are utilized to flatten out and relieve fluctuation issues. To prevent the need for larger storage systems and to prolong their operational life through controlled charging and discharging, a method of control for BESS charging level regulation is necessary. This study presents a solar-wind power and battery state of charge (SoC) control technique using a hydrogen electrolyzer (HE) fuel cell unit. An Intelligent Model Predictive Controller (IMPC) has been developed that utilizes a neural network (NN) plant model for predictive minimization instead of using a mathematical plant model for dynamic management of the HE output, which allows for flattened PV-wind power supply with regulated BESS charging and discharging. The IMPC accepts as inputs the intermittent renewable power and different battery attributes and intelligently manages the HE production while staying within the imposed restrictions. The NN, as opposed to a mathematical plant model, captures the dynamics of the plant with exceptionally high accuracy. Moreover, the NN plant model performance increases as the data gathered from the actual system increases. The neural network model resolves concerns with the MPC model's mathematical intricacy that occurs as the plant becomes more complicated. According to simulation results, the IMPC greatly reduces PV-wind power fluctuations, for solar power, the IMPC reduces the peak battery SoC by 26.7% and compared to FLC, the peak SoC is reduced by 11.2%. Similarly, for wind power, the peak SoC is reduced by 7.3% and is decreased 3.2% more than the FLC. To demonstrate the power smoothing effectiveness of the IMPC, the peak solar power ramp rate is reduced by 30.3% and the peak wind power ramp rate is reduced by 91.3%. The presented method encourages green hydrogen usage in the electrical sector for supplying firmed solar and wind power.

Automated summary

This study introduces a solar-wind power and battery state of charge control technique using a hydrogen electrolyzer fuel cell unit. By utilizing an Intelligent Model Predictive Controller and a neural network plant model, it achieves stable PV-wind power supply with regulated battery energy storage system charging and discharging.