摘要：Hydrogen network synthesis (HNS) in refineries serves as an effective approach to enhance hydrogen utilization efficiency, optimize network topology, and reduce operating costs. Conventional methods typically rely on the fixed hydrogen source and sink parameters, which deviates from the refinery reality. To address this problem, this work develops a data-driven optimization framework that hybrids rigorous simulation and machine learning to establish a coupling hydrogen sink-source model with hydroprocessing mechanisms and product specifications. The proposed framework comprehensively incorporates reaction-separation-compression systems in HNS. A mixed-integer nonlinear programming model is formulated for optimal synthesis of hydrogen networks. The results of systematic case studies reveal that implementing the coupling sink-source model expands the optimization dimensions, thereby reducing the pure hydrogen consumption in hydroprocessing units and decreasing the total annualized cost (TAC) by 9.35 %similar to 10.00 %. The consideration of compressor sharing contributes to a slight TAC reduction due to its low-cost proportion, but it enables additional cost savings through topology reconfiguration. The synergistic optimization of hydrogen networks and hydroprocessing units is achieved in this work.

关键字：Hydrogen network synthesis; Hydroprocessing mechanisms; Coupling sink-source model; Machine learning; Mixed-integer nonlinear programming

ISSN号：0009-2509

卷、期、页：卷: 321子辑: B

发表日期：2026-02-01

期刊分区（SCI为中科院分区）：二区

收录情况：SCI（科学引文索引印刷版）,EI（工程索引）,SCIE（科学引文索引网络版）

发表期刊名称：CHEMICAL ENGINEERING SCIENCE

参与作者：Zhou, Yingqian,Yang, Minbo

第一作者：王彧斐

论文类型：期刊论文

论文概要：Zhou, Yingqian,Yang, Minbo,王彧斐，Modeling and optimization of refinery hydrogen management system considering hydroprocessing mechanisms and constraints，CHEMICAL ENGINEERING SCIENCE，2026，卷: 321子辑: B

论文题目：Modeling and optimization of refinery hydrogen management system considering hydroprocessing mechanisms and constraints