基于功率因数约束的多电堆调峰负荷控制框架
A Multi-stack Power-to-Hydrogen Load Control Framework for the Power Factor-Constrained Integration in Volatile Peak Shaving Conditions
摘要：
多电堆组成的大规模发电制氢（P2H）系统是电力系统潜在的强大调峰资源。然而，由于在连接电网侧的性能与固有的运行控制的研究差距，P2H负荷的连续运行是有限的PF评估时，融入电网的波动条件下。本文首先填补了空白，提出了一个典型的功率变换器接口拓扑的P2H负载的有功功率和无功功率的分析模型。在此基础上，捕捉多堆P2H负载的全工况PF特性作为统一的电流和温度控制变量的函数。然后，PF约束的多时标控制框架的构建，以综合评估柔性，PF，生产和安全。为保证控制精度和易处理性，提出了基于模型的小时前鲁棒模型预测控制器和基于规则的实时增量校正算法的两级连接。案例研究验证了一个内在的控制PF和生产之间的权衡，从而在一个不平等的分配策略相比，传统的面向生产的控制。验证了扩展PF和安全维度的意义，以提高灵活性。在此基础上，归纳出了集群级低、中、高负荷下的5种典型运行模式，可供工业应用。
英文摘要：
Large-scale power-to-hydrogen (P2H) systems formed by multi-stack are potentially powerful peak-shaving resources of power systems. However, due to the research gap in connecting the grid-side performance with the inherent operation control, the continuous operation of P2H loads is limited by the PF assessment under volatile conditions when integrating into the grid. This paper first fills the gap in proposing the analytical models of active and reactive power of P2H loads with a typical power converter interface topology. On this basis, the all-condition PF characteristics of multi-stack P2H loads are captured as functions of unified current and temperature control variables. Then, a PF-constrained multi-timescale control framework is constructed to evaluate flexibility, PF, production, and security comprehensively. A two-level nexus, including a model-based hour-ahead robust model predictive controller and a rule-based real-time increment correction algorithm, is proposed to guarantee the control accuracy and tractability. Case studies verify an intrinsic control tradeoff between PF and production, resulting in an unequal-split allocation strategy compared to the traditional production-oriented control. The significance of the extended PF and security dimensions is verified to improve the flexibility. Furthermore, five typical operating modes respectively corresponding to low, medium, and high load levels at the cluster level are concluded for industrial application.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
