Objective Function Research Articles

Multi-objective geometric optimization of protrusion, droplet ribs, and inclined upper plate of a slit jet impingement heat sink to enhance its thermal performance

To meet the needs of heat transfer and temperature uniformity in the heat dissipation of high-heat-flux electronic devices, a novel slit jet impingement heat sink with a protrusion in the stagnation region, droplet ribs in the wall jet region, and an inclined upper-plate (PDI-SJIHS) is proposed, the coolant is Al2O3-H2O nanofluid. The influences of the structural parameters of the above three components and the Reynolds number on the heat transfer and flow of the PDI-SJIHS are studied numerically. Using the comprehensive heat transfer and temperature standard deviation as objective functions, a multi-objective optimization of PDI-SJIHS is conducted using non-dominated sorting genetic algorithm-Ⅱ (NSGA-Ⅱ). The results indicate that the combination of the three components enhances the thermal performance of the PDI-SJIHS, compared with the slit jet impingement heat sink with only one of the above components. Rising the Reynolds number can enhance the thermal characteristics of the PDI-SJIHS. For a Reynolds number of 8000 and nanofluids with a 3% volume fraction, the average convective heat transfer coefficient, friction coefficient, and comprehensive heat transfer coefficient of the optimized PDI-SJIHS are increased by 182%, 153%, and 108%, respectively, and the standard deviation of temperature is 91% less than that of F-SJIHS.

Evolutionary optimization via swarming dynamics on products of spheres and rotation groups

We propose novel gradient-free algorithms for optimization problems where the objective functions are defined on products of spheres or rotation groups. Optimization problems of this kind are common in robotics and aeronautics where learning rotations and orientations in space is one of the core tasks. Moreover, in many cases it is required to find several mutually dependent orientations or several coupled rotations, making the optimization problem much more demanding. Our approach is based on recently introduced families of probability distributions, as well as on trainable swarms on spheres and rotation groups. The underlying idea is that models and architectures in robotics and machine learning are to a great extent imposed by geometry of the data. The proposed approach is flexible and can be adapted to setups with sequential (temporal) data. In order to make our methods clearer, a number of illustrative problems are introduced and solved using the proposed methods.

Product tracing or component tracing? Blockchain adoption in a two-echelon supply chain management

Consumer awareness of product authenticity and carbon footprint tracing are among the most remarkable reasons for adopting blockchain in the supply chain in today’s world. However, the research literature has not yet examined specific ways to adopt blockchain in the supply chain. This study aims to develop a decision support tool to deal with the adoption of blockchain technology to design a two-echelon supply chain. In this regard, four specific cases for integrating supply chain and blockchain are developed based on types of tracing and block generation authority. In product tracing, green products are investigated throughout the supply chain, while in component tracing, green products are examined between the components of the supply chain. As it is necessary to record and verify the supply chain information by authorities in the blockchain network, in this work, such authorities are taken into account for both links and members of the supply chain. As far as we know, this is the first attempt to classify the various methods of adopting blockchain in Green Supply Chain Management (GSCM) and propose mathematical optimization models related to them. In this line, four Mixed-Integer Linear Programming (MILP) models with the aim of minimizing the costs related to the physical supply chain and blockchain deployment are developed for the integration of the supply chain with blockchain technology. They are treated by the Branch and Efficiency (B&E) algorithm and Simultaneous Data Envelopment Analysis (SDEA) model considering common (cost and service) and innovative (blockchain) criteria. The results showed that link-based and component tracing models are cost-effective. In addition, the cost objective function of green product tracing is more sensitive to the number of blocks than that of component tracing. Eventually, the study provides great opportunities for decision-makers and managers to understand how to adopt blockchain in terms of supply chain network characteristics, cost, transparency, and service.

GDFold2: A fast and parallelizable protein folding environment with freely defined objective functions.

An important step of mainstream protein structure prediction is to model the 3D protein structure based on the predicted 2D inter-residue geometric information. This folding step has been integrated into a unified neural network to allow end-to-end training in state-of-the-art methods like AlphaFold2, but is separately implemented using the Rosetta folding environment in some traditional methods like trRosetta. Despite the inferiority in prediction accuracy, the conventional approach allows for the sampling of various protein conformations compatible with the predicted geometric constraints, partially capturing the dynamic information. Here, we propose GDFold2, a novel protein folding environment, to address the limitations of Rosetta. On the one hand, GDFold2 is highly computationally efficient, capable of accomplishing multiple folding processes in parallel within the time scale of minutes for generic proteins. On the other hand, GDFold2 supports freely defined objective functions to fulfill diversified optimization requirements. Moreover, we propose a quality assessment (QA) model to provide reliable prediction on the quality of protein structures folded by GDFold2, thus substantially simplifying the selection of structural models. GDFold2 and the QA model could be combined to investigate the transition path between protein conformational states, and the online server is available at https://structpred.life.tsinghua.edu.cn/server_gdfold2.html.

Impact analysis of uncertainty in thermal resistor-capacitor models on model predictive control performance

Model Predictive Control (MPC) is extensively utilized for optimal control in building systems. Despite substantial research being dedicated to exploring the impact of uncertainties in external and internal disturbances on the performance of MPC, the existing studies neglect the potential impact of uncertainties in model parameter identification on control performance. To address this gap, this study quantifies the impact of model uncertainty on MPC performance through a test case in a virtual environment. Various levels of uncertainties for parameters R and C are artificially introduced to assess the MPC performance. The causes of the impact of model uncertainty on control performance are further explored through analysis. We select a first-order RC model to modelling building thermal dynamics. MPC is employed to optimize the heat pump signal with the goal of minimizing the energy cost while maintaining thermal comfort. The simulation results demonstrate that a negative deviation in model parameter identification has a more pronounced impact on MPC performance than a positive deviation, which has a negligible effect on MPC control performance. Deviations in parameters from their true values affect both heat losses from the zone and thermal capacity, thereby influencing the estimated temperature by the RC model. Consequently, these factors, in turn, affect the system’s control decisions, leading to variations in the objective function values. This study can provide an insight into the relationship of model parameters uncertainties and MPC performance and facilitate the practical application of MPC in buildings.

A sustainable and resilient humanitarian relief chain network design for distributing assembled relief items dynamically considering perishability, under disruption

Purpose Due to mitigate against natural disasters like earthquake and to distribute relief items, designing humanitarian relief chain networks is an attentional issue. Agile and efficient distribution of relief items after occurring a disaster is significant, especially when some of the relief items are perishable. Therefore, the purpose of this paper is to create a resilient and integrated decision-making structure to distribute relief items at demand points, considering two dimensions of sustainability, under disruption. Design/methodology/approach This study developed a mixed-integer nonlinear mathematical model to handle the pre- and post-disaster planning when a disaster occurs. The represented model has two objective functions: minimizing weighted unmet demand and total costs. Therefore, to convert this multi-objective problem into a single objective one, the e-constraint method was applied. Findings The main results showed that considering some resilience strategies has a significant effect in reducing the weighted amount of unmet demand and saves the total costs. More precisely, considering resilience strategies results in a 60% reduction in total unmet demand and 11% reduction in total pre-positioning costs. On the other hand, reducing the maximum response time with applying resilience strategies is another achievement of the present study. For these reasons, the use of these strategies can reduce people’s pain and suffer from natural disasters. In general, the application and effectiveness of sustainability dimensions and resilience strategies in the introduced humanitarian relief chain network were analyzed. Practical implications To verify the applicability of this study, this model is applied on a probable real-life case study in Tehran. Finally, some managerial insights are discussed to help humanitarian organizations, managers and stakeholders to make better decisions to reduce negative effects of natural disasters. Originality/value This paper introduced a two-stage stochastic mathematical model for designing a resilient humanitarian relief chain network under disruption, at pre- and post-disaster stages. Also, economic and social dimensions of sustainability are considered in this study. Moreover, assembling perishable and im-perishable relief items as relief kits, dynamically is a main contribution of this research.

Directed capacity-preserving subgraphs: hardness and exact polynomial algorithms

We introduce and discuss the Minimum Capacity-Preserving Subgraph (MCPS) problem: given a directed graph with edge capacities cap and a retention ratio α∈(0,1), find the smallest subgraph that, for each pair of vertices (u, v), preserves at least a fraction α of a maximum u-v-flow’s value. This problem originates from the practical setting of reducing the power consumption in a computer network: it models turning off as many links as possible, while retaining the ability to transmit at least α times the traffic compared to the original network. First we prove that MCPS is NP-hard already on a restricted set of directed acyclic graphs (DAGs) with unit edge capacities. Our reduction also shows that a closely related problem (which only considers the arguably most complicated core of the problem in the objective function) is NP-hard to approximate within a sublogarithmic factor already on DAGs. In terms of positive results, we present two algorithms that solve MCPS optimally on directed series-parallel graphs (DSPs): a simple linear-time algorithm for the special case of unit edge capacities and a cubic-time dynamic programming algorithm for the general case of non-uniform edge capacities. Further, we introduce the family of laminar series-parallel graphs (LSPs), a generalization of DSPs that also includes cyclic and very dense graphs. Their properties allow us to solve MCPS on LSPs by employing our DSP-algorithms as subroutines. In addition, we give a separate quadratic-time algorithm for MCPS on LSPs with unit edge capacities that also yields straightforward quadratic time algorithms for several related problems such as Minimum Equivalent Digraph and Directed Hamiltonian Cycle on LSPs.

Optimal Sizing and Techno-economic Analysis of Combined Solar Wind Power System, Fuel Cell and Tidal Turbines Using Meta-heuristic Algorithms: A Case Study of Lavan Island

Combined renewable energy sources (RESs) are emerging as a competitive alternative to conventional energy production facilities due to their sustainability and zero-emission characteristics. However, determining the optimal system size is complicated by two major challenges: the cost of energy (COE) and the intermittent nature of RESs. This study introduces a novel mathematical approach to optimize the sizing of photovoltaic (PV), wind, hydrogen, battery, and fuel cell systems with electrolyzers, specifically tailored for the remote area of Lavan Island. The proposed method aims to deliver electricity without reliance on the traditional electricity distribution grid, while offering a scalable solution applicable to other geographical regions. The primary objective is to achieve cost-effective electricity generation while ensuring a reliable energy supply through the evaluation of system reliability indices. A fuzzy logic system is employed to minimize the costs of a hybrid system incorporating hydroelectric, wind, solar, and battery technologies, while simultaneously calculating two key reliability metrics: the Loss of Power Supply Probability (LPSP) and the Dump Energy Probability (DEP). To optimize the objective function, this study applies three advanced algorithms: the Shuffled Frog Leaping Algorithm (SFLA), the Grasshopper Optimization Algorithm (GOA), and the Honey Badger Algorithm (HBA). These algorithms are used to determine the global optimum, with comparative analyses conducted to highlight the performance of the proposed approach. The results are evaluated based on statistical metrics, including consistency, execution time, convergence speed, and the minimization of the objective function. The findings demonstrate the superiority and the reliability of the proposed method over alternative approaches, paving the way for cost-efficient and sustainable energy solutions in isolated regions.

Classification of CT scan and X-ray dataset based on deep learning and particle swarm optimization.

In 2019, the novel coronavirus swept the world, exposing the monitoring and early warning problems of the medical system. Computer-aided diagnosis models based on deep learning have good universality and can well alleviate these problems. However, traditional image processing methods may lead to high false positive rates, which is unacceptable in disease monitoring and early warning. This paper proposes a low false positive rate disease detection method based on COVID-19 lung images and establishes a two-stage optimization model. In the first stage, the model is trained using classical gradient descent, and relevant features are extracted; in the second stage, an objective function that minimizes the false positive rate is constructed to obtain a network model with high accuracy and low false positive rate. Therefore, the proposed method has the potential to effectively classify medical images. The proposed model was verified using a public COVID-19 radiology dataset and a public COVID-19 lung CT scan dataset. The results show that the model has made significant progress, with the false positive rate reduced to 11.3% and 7.5%, and the area under the ROC curve increased to 92.8% and 97.01%.

Research on Optimized Multi‐Objective FCS‐MPC Without Weighting Factors of NPC Three‐Level Inverter

ABSTRACTTo address a series of issues in the application of finite control set model predictive control (FCS‐MPC) strategy for neutral point clamped (NPC) three‐level inverters, such as large amount of rolling optimization calculation, poor midpoint voltage balance effect, high common mode voltage, and complex design of weighting coefficients, this paper proposes an optimized multi‐objective FCS‐MPC strategy without weighting factors. First, the method applies the idea of space vector region division, to obtain the optimal region based on Lyapunov function, and combines with delay compensation, reducing the system's optimization time and enhancing the control precision of the system. Second, aiming at the issue of multi‐objective control for inverters, by constructing the objective function to control tracking reference current, midpoint voltage balance, and reduce common mode voltage, a hierarchical optimization control method is adopted. This approach eliminates the need to design complex weighting coefficients constrained in the objective function as constraints, thereby improving the accuracy of predicted currents. Finally, the optimized multi‐objective FCS‐MPC control strategy is validated through simulation and experimentation to achieve a better balance of midpoint voltage and reduce common mode voltage and harmonic content, thereby enhancing the control precision and dynamic/static performance of the system.

Parallel Bayesian optimization of free-electron lasers based on laser wakefield accelerators

Abstract In this paper, the parallel Bayesian optimization algorithm is investigated for the simulation optimization of compact free electron lasers (FELs) driven by laser wakefield accelerators (LWFAs). The radiation energy serves as the objective function, which can be optimized to 25 μJ after several tens of iterations, initiating from random input samplings in soft x-ray regimes. The analysis of the FEL gain process has revealed that the localized properties of the electron beam (e beam) can be concurrently optimized, ensuring a high efficiency of energy extraction from the e beam. Our proposed scheme not only presents a viable design for compact FELs driven by LWFAs in the soft x-ray regime utilizing the start-to-end model, but also highlights the great potential of parallel Bayesian optimization algorithms for tackling the challenges associated with costly and complex black-box optimization problems.

An Improved Marriage in Honey-Bee Optimization Algorithm for Minimizing Earliness/Tardiness Penalties in Single-Machine Scheduling with a Restrictive Common Due Date

This study evaluates the efficiency of a swarm intelligence algorithm called marriage in honey-bee optimization (MBO) in solving the single-machine weighted earliness/tardiness problem, a type of NP-hard combinatorial optimization problem. The goal is to find the optimal sequence for completing a set of tasks on a single machine, minimizing the total penalty incurred for tasks being completed too early or too late compared to their deadlines. To achieve this goal, the study adapts the MBO metaheuristic by introducing modifications to optimize the objective function and produce high-quality solutions within reasonable execution times. The novelty of this work lies in the application of MBO to the single-machine weighted earliness/tardiness problem, an approach previously unexplored in this context. MBO was evaluated using the test problem set from Biskup and Feldmann. It achieved an average improvement of 1.03% across 280 problems, surpassing upper bounds in 141 cases (50.35%) and matching or exceeding them in 193 cases (68.93%). In the most constrained problems (h = 0.2 and h = 0.4), the method achieved an average improvement of 3.77%, while for h = 0.6 and h = 0.8, the average error was 1.72%. Compared to other metaheuristics, MBO demonstrated competitiveness, with a maximum error of 1.12%. Overall, MBO exhibited strong competitiveness, delivering significant improvements and high efficiency in the problems studied.

Distributed planning method for detection array of UAV formation with bearing-only sensor network

Abstract Unmanned aerial vehicle (UAV) formations for bearing-only passive detection are increasingly important in modern military confrontations, and the array of the formation is one of the decisive factors affecting the detection accuracy of the system. How to plan the optimal geometric array in bearing-only detection is a complex nondeterministic polynomial problem, and this paper proposed the distributed stochastic subgradient projection algorithm (DSSPA) with layered constraints to solve this challenge. Firstly, based on the constraints of safe flight altitude and fixed baseline, the UAV formation is layered, and the system model for bearing-only cooperative localisation is constructed and analysed. Then, the calculation formula for geometric dilution of precision (GDOP) in the observation plane is provided, this nonlinear objective function is appropriately simplified to obtain its quadratic form, ensuring that it can be adapted and used efficiently in the system model. Finally, the proposed distributed stochastic subgradient projection algorithm (DSSPA) combines the idea of stochastic gradient descent with the projection method. By performing a projection operation on each feasible solution, it ensures that the updated parameters can satisfy the constraints while efficiently solving the convex optimisation problem of array planning. In addition to theoretical proof, this paper also conducts three simulation experiments of different scales, validating the effectiveness and superiority of the proposed method for bearing-only detection array planning in UAV formations. This research provides essential guidance and technical reference for the deployment of UAV formations and path planning of detection platforms.

A Globalization of L-BFGS and the Barzilai–Borwein Method for Nonconvex Unconstrained Optimization

We present a modified limited memory BFGS (L-BFGS) method that converges globally and linearly for nonconvex objective functions. Its distinguishing feature is that it turns into L-BFGS if the iterates cluster at a point near which the objective is strongly convex with Lipschitz gradients, thereby inheriting the outstanding effectiveness of the classical method. These strong convergence guarantees are enabled by a novel form of cautious updating, where, among others, it is decided anew in each iteration which of the stored pairs are used for updating and which ones are skipped. In particular, this yields the first modification of cautious updating for which all cluster points are stationary while the spectrum of the L-BFGS operator is not permanently restricted, and this holds without Lipschitz continuity of the gradient. In fact, for Wolfe–Powell line searches we show that continuity of the gradient is sufficient for global convergence, which extends to other descent methods. Since we allow the memory size to be zero in the globalized L-BFGS method, we also obtain a new globalization of the Barzilai–Borwein spectral gradient (BB) method. The convergence analysis is developed in Hilbert space under comparably weak assumptions and covers Armijo and Wolfe–Powell line searches. We illustrate the theoretical findings with numerical experiments. The experiments indicate that if one of the parameters of the cautious updating is chosen sufficiently small, then the modified method agrees entirely with L-BFGS/BB. We also discuss this in the theoretical part. An implementation of the new method is available on arXiv.

Robust optimal model of green vendor-managed inventory for carbon emission reduction and supply chain visibility

Green supply chain management (SCM) and the green vendor-managed inventory (VMI) strategy are increasingly becoming the focus of the academic world. This paper innovatively investigates the green VMI problems in a three-tier supply chain network, especially incorporating multiple constraints such as supply chain visibility (SCV), quality constraints, carbon emission allowance trading mechanism and green technology investment. To solve this multi-constraint problem, a robust multi-objective optimisation model is constructed in this paper to handle the demand uncertainty. Example values verify the model’s performance, and the sensitivity analysis of critical parameters is carried out. The results show that the robust model can effectively realise the optimal solution of the whole multi-objective system by balancing the optimal value of a single objective function under the multi-objective model. In addition, maintaining a low defect rate and delivery delay rate is crucial for efficient supply chain operation. Furthermore, SCV and supply chain green efficiency need to be balanced, and the pursuit of high SCV will inevitably increase the carbon emission cost of the supply chain system. However, the results also show that investment in green technology R&D can help to improve this situation and realise the effective reduction of carbon emissions while increasing SCV.

Online Kernel-Based Mode Learning

The presence of big data, characterized by exceptionally large sample size, often brings the challenge of outliers and data distributions that exhibit heavy tails. An online learning estimation that incorporates anti-outlier capabilities while not relying on historical data is therefore urgently required to achieve robust and efficient estimators. In this paper, we introduce an innovative online learning approach based on a mode kernel-based objective function, specifically designed to address outliers and heavy-tailed distributions in the context of big data. The developed approach leverages mode regression within an online learning framework that operates on data subsets, which enables the continuous updating of historical data using pertinent information extracted from a new data subset. By amalgamating the asymptotic distribution functions generated by local mode estimators from all data subsets, the newly suggested estimator is efficiently updated by minimizing a weighted least squares-type loss function. To facilitate this process, we suggest a modified mode expectation-maximization algorithm for numerical optimization, ensuring both storage friendliness and computational efficiency. We demonstrate that the resulting estimator is asymptotically equivalent to the mode estimator calculated using the entire dataset, provided that the covariates in each data subset are homogeneous and the errors are homoskedastic. Monte Carlo simulations and an empirical study are presented to illustrate the finite sample performance of the proposed estimator. Supplemental materials for this paper are available online.

Privacy-Preserving Incentive Allocation for Fair and Resilient Data Sharing in Resource-Constrained Edge Computing Networks

Efficient and secure data sharing is paramount for advancing modern digital ecosystems, especially within edge computing environments characterized by resource-constrained nodes and dynamic network topologies. In such settings, privacy preservation, computational efficiency, and system resilience are critical for user engagement and overall system performance. However, existing approaches face three primary challenges: (i) limited optimization of privacy protection and absence of dynamic privacy budget scheduling for resource-constrained scenarios, (ii) static incentive mechanisms that overlook individual differences in data quality and resource consumption, and (iii) inadequate strategies to ensure resilience in environments with limited resources and unstable networks. This paper introduces the Federated Learning-based Dynamic Incentive Allocation Framework (FL-DIAF) to address these issues. FL-DIAF integrates differential privacy into the federated learning paradigm deployed on edge nodes, enabling collaborative model training that safeguards individual data privacy while maintaining computational efficiency and system resilience. Additionally, the framework employs a Shapley value-based dynamic incentive allocation model to ensure equitable and transparent distribution of incentives by accurately quantifying each participant’s contribution within an elastic edge computing infrastructure. Comprehensive experimental evaluations on diverse datasets demonstrate that FL-DIAF achieves a 9.573% reduction in the objective function value under typical conditions and attains a 100% task completion rate across all tested resilient edge scenarios.

The Tolerance Interval Optimization of Cable Forces During the Construction Phase of Cable-Stayed Bridges Based on Hybrid Intelligent Algorithms

To investigate the controllability of sensitive cable forces during the construction phase of cable-stayed bridges, a novel optimization method is proposed, based on BP neural networks, which combines Gaussian process prediction with a simulated annealing-optimized particle swarm algorithm to determine the tolerance intervals of construction cable forces. Based on the analysis results of multiple linear regression, the variables for optimization are identified, and a mapping relationship between the sensitive cable forces and displacement values is established using a BP neural network. Subsequently, a Gaussian process model is constructed to delineate the relationship between cable forces and reliability, with a focus on the reliability of displacements during the construction phase of the cross-section, specifically targeting sensitive cable forces. Finally, a combination of the simulated annealing algorithm and the particle swarm algorithm is employed to optimize the tolerance intervals of the cable forces. To validate the effectiveness of the proposed optimization method, a case study is conducted on the tolerance interval optimization of cable forces using a three-tower steel box girder cable-stayed bridge. In this study, the construction cable forces are treated as optimization variables, while the reliability of displacements at both the main girder section and the tower’s top section serve as the optimization objectives and constraint conditions. Under the premise of ensuring structural reliability, the accurate tolerance range for the stay cable forces during the construction phase of the cable-stayed bridge is obtained. The results indicate that the traditional PSO algorithm stabilizes after 26 iterations, whereas the hybrid intelligent algorithm reaches stability after just 13 iterations. In addition, the hybrid algorithm shows a significant increase in the objective function value during early iterations, demonstrating stronger optimization capability. This indicates that the optimization method exhibits better convergence and superior global optimization capability. It effectively improves the compatibility and controllability of the cable-stayed bridge construction process while simplifying the computational process.

Optimal design of multivariate accelerated degradation test based on game theory

ABSTRACT Optimal design of the accelerated degradation test plays an important role in improving the reliability assessment accuracy of products. The commonly used optimal design methods of accelerated degradation tests are mainly based on a single performance degradation characteristic. When considering multiple-performance degradation characteristics, there is a problem that the resulting test plans cannot be unified. To solve the problem, a multivariate accelerated degradation test optimization method is presented based on game theory in this work. The accelerated degradation models are established with the Wiener process as the degradation model assumption, and the correlation is described by the Copula function. The optimization model is established with the objective function of maximizing the Fisher information matrix determinant for each degradation model. Based on game theory, the optimization model solution problem is transformed into a cooperative game problem, and the optimal test plan is obtained by maximizing the benefit function. Finally, the effectiveness of the proposed method is verified by an example of O-rings.

Integrated Planning Model for Optimizing Investments in Hydrogen Infrastructure Deployment: A Case Study of Yangtze River Delta Region, China

ABSTRACTThe hydrogen industry is of great significance for global energy system transition and decarbonization; thus, the holistic planning of hydrogen infrastructure from a supply chain level is necessary. To this end, first, the framework and methods in the hydrogen infrastructure deployment research were investigated systematically. Second, a hydrogen supply chain design (HSCD) model and a hydrogen refueling station location (HRSL) model were constructed. In view of the limitations of the two models, this paper reformulated the HRSL model through alternating the objective function and constraint, then proposed an integrated planning model using the ‐constraint method, which relaxed the assumption condition and expanded the decision‐making boundary of the original model, realizing the reasonable equilibrium of multiple segments. Finally, a case study was conducted using the Yangtze River Delta region as an example, and the main findings were as follows: (1) To meet the hydrogen demand of 25494t/d, the Yangtze River Delta region needs to build 94 coal gasification hydrogen plants, storage facilities with capacity of 12747t, transmission modes with capacity of 2333t/d, and 150 hydrogen refueling stations. (2) Railways have more advantages in large capacity and long‐distance hydrogen transmission. (3) Road segments with higher traffic flow often have a higher capture proportion. (4) The investment cost accounts for the majority of hydrogen production plant construction, which is 43.80%.