高级检索

    基于IGDT的可再生能源制氢系统四对一碱性电解集群优化调度

    Optimal scheduling of 4-in-1 alkaline electrolyzers in renewable power-to-hydrogen systems based on Information Gap Decision Theory

    • 摘要: 规模化可再生能源发电制氢可有效消纳风光资源,为下游产业供给绿氢,助力化工、交通等行业减碳。为减少设备以及占地成本,部分工程采用多台碱性电解槽共用一套辅机系统的“多对一”工艺构型。但受限于辅机系统耦合,多对一系统中多台电解槽启停以及运行功率范围相互影响,相较一对一构型灵活性降低。为解决多对一电解集群应对可再生能源波动性与随机性难题,提出基于信息间隙决策理论(information gap decision theory, IGDT)的四对一碱性电解集群优化调度方法。首先,对电解槽的功率、产氢量及启停组合进行建模。其次,基于工程中四对一系统运行逻辑,刻画电解槽间耦合约束。然后,通过鲁棒IGDT建模可再生能源的不确定性,提出电制氢系统优化调度模型,并构造混合整数二阶锥规划予以求解。基于内蒙古某实际工程的算例分析表明,所提调度模型最多可提升收益2.3%,有效应对风光出力不确定性。在收益偏差因子为0.15时,可接受的风光预测误差达20%,展现出较好的鲁棒性。最后,定量讨论了多对一制氢集群与一对一系统在消纳可再生能源方面的差异。

       

      Abstract: Large-scale renewable power-to-hydrogen (ReP2H) systems enable efficient utilization of wind and solar resources while supplying green hydrogen to downstream industries, thereby decarbonizing chemical and transportation sectors. To reduce equipment and land-use costs, some projects adopt N-in-1 configurations, where multiple stacks share a common balance-of-plant (BoP) system. However, due to the coupling of gas–liquid circulation processes, the startup and shutdown of individual stacks become interdependent, leading to reduced operational flexibility compared with conventional one-in-one configurations. To address the challenges of N-in-1 electrolyzers under the variability and uncertainty of renewable generation, this paper proposes a scheduling method for multiple 4-in-1 electrolyzers based on information gap decision theory (IGDT). First, the power consumption, hydrogen production, and on-off switching of the stacks are explicitly modeled. Second, the inter-stack couplings are characterized according to the operational logic of practical 4-in-1 systems. Finally, renewable uncertainty is handled using a robust IGDT framework, and the resulting scheduling problem is formulated as a mixed-integer second-order cone programming (MISOCP) model. Case studies demonstrate that the proposed method increases revenue by up to 2.3% while effectively handling renewable uncertainty. With a revenue deviation factor of 0.15, the system can tolerate forecast errors of up to 20%, indicating strong robustness. Furthermore, the performance differences between N-in-1 and one-in-one configurations in accommodating renewable energy are quantitatively analyzed.

       

    /

    返回文章
    返回