Smart Operation Optimization Methods for Hydrogen-Based Energy Systems Under Uncertainties

The operation optimization of hydrogen-based energy systems (HESs) contributes to the reduction of operational cost and carbon emission. In existing works, operation optimization methods either require prior information of uncertain parameters or do not consider the nonlinear characteristics of electrolyzers and fuel cells. To overcome the drawbacks in existing works, we investigate an expected operational cost minimization problem related to HESs with the consideration of nonlinear characteristics and propose an online operation algorithm to minimize the total operational cost of HESs based on proximal policy optimization and expert knowledge. Moreover, the proposed algorithm does not require any prior knowledge of uncertain parameters. Simulation results show that the proposed algorithm can reduce the total operational cost of HESs by 2.7%-20.5% and carbon emission by 15.6%-44.3% compared with the rule-based scheme.