Online Operation Optimization for Hydrogen-Based Building Energy Systems Under Uncertainties

Hydrogen-based energy systems have attracted widespread attention due to their capabilities of promoting renewable penetration and cutting carbon emissions. In this article, we study an operation optimization problem of a hydrogen-based building energy system. Since there are multi-source uncertainties related to electricity price, hydrogen price, electrical load, thermal load, and renewable generation output, spatially and temporally coupled constraints associated with power balance, heat balance, and energy storage systems, and nonlinear constraints, it is difficult to solve the considered optimization problem. To deal with the above difficulties, this paper proposes a low-complexity online operation optimization algorithm. Furthermore, we theoretically analyze the algorithmic feasibility and performance guarantees. Extensive simulations verify that the proposed algorithm can achieve near-optimal performance and reduce operation cost by 2.81%-7.74% compared with five baselines, including model predictive control and deep reinforcement learning.