Two-stage trigger dispatch strategy for hydrogen-electricity integrated station based on hybrid energy storage under response willingness uncertainty
Yanjia Wang, Xi Cheng, Mohannad Alhazmi, Shilin Chen, Min Xu, Da Xie, Xitian Wang
Abstract
This paper addresses the coordinated optimization of power and hydrogen systems within a multi-energy system (MES) framework by proposing a two-stage trigger dispatch strategy for hydrogen-electricity integrated stations considering uncertainty in response willingness. The study focuses on two main aspects: the design of the hydrogen-electricity integrated station and the two-stage optimal dispatch strategy. First, an intelligent integrated station is developed by incorporating hydrogen production , hydrogen storage , fuel cell (FC), hydrogen refueling for hydrogen vehicles (HVs), and electric vehicles (EVs) charging and discharging. HVs, EVs, and hydrogen storage form a hybrid energy storage (HES) that interacts with the grid through an integrated station. This integration facilitates the coordination between the power grid and the hydrogen system, enhancing the overall flexibility and efficiency of the MES. In the two-stage dispatch strategy, the first stage employs a scheduling optimization allocation approach that comprehensively considers factors such as the operational efficiency of each integrated station, the response willingness of vehicle users, and the states of both electrical and hydrogen storage. A convolutional neural network (CNN) is introduced to accurately model the response willingness of EVs around charging stations, thereby improving the precision of power distribution. The second stage introduces a triggering dispatch mechanism that mitigates the impact of uncertainties in the response from HVs and EVs through continuous interactive scheduling between the integrated station and the vehicles. Additionally, an economic analysis of the FC response , including the recovery process, is incorporated to enhance the economic efficiency of the multi-energy coordinated response. The results show that, compared to existing research, the proposed method reduces the total dispatch cost by 10.74 %.