Journal of Computational Science最新文献_第4页

Graph anomaly detection via local contrastive learning and attribute reconstruction 基于局部对比学习和属性重构的图异常检测

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-02-01 Epub Date: 2026-01-02 DOI: 10.1016/j.jocs.2025.102781

Xinyu Zhang , Yanfen Li

{"title":"Graph anomaly detection via local contrastive learning and attribute reconstruction","authors":"Xinyu Zhang , Yanfen Li","doi":"10.1016/j.jocs.2025.102781","DOIUrl":"10.1016/j.jocs.2025.102781","url":null,"abstract":"<div><div>Graph anomaly detection (GAD) plays a vital role in identifying abnormal nodes within graph data, with applications in social networks, fraud detection, and cybersecurity. However, existing approaches face challenges such as anomaly overfitting and the homophily trap, which hinder accurate detection. Anomaly overfitting occurs when models are excessively influenced by anomalies during training, while the homophily trap arises from the assumption that connected nodes share similar features, which can mislead the model. To address these challenges, we propose Contrastive Local Anomaly-Aware Reconstruction Embedding (CLARE), a novel method designed to overcome these limitations. CLARE employs a contrastive sampling strategy to construct a local reference distribution, thus enhancing the learning of normal patterns. It integrates attribute and structural reconstruction to detect anomalies based on reconstruction errors. A key innovation is the neighbor and central node reconstruction mechanism, which improves detection accuracy by incorporating second-order neighbor information. Experimental results demonstrate that CLARE outperforms existing methods, offering robust and scalable anomaly detection for complex graph data.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"94 ","pages":"Article 102781"},"PeriodicalIF":3.7,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146022642","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A liquid neural network with physical evolution for variable continuous time series prediction 一种具有物理演化的液体神经网络用于变量连续时间序列预测

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-02-01 Epub Date: 2025-11-27 DOI: 10.1016/j.jocs.2025.102757

Kui Qian, Yue Deng, Zhengyan Li, Xiulan Wen

{"title":"A liquid neural network with physical evolution for variable continuous time series prediction","authors":"Kui Qian, Yue Deng, Zhengyan Li, Xiulan Wen","doi":"10.1016/j.jocs.2025.102757","DOIUrl":"10.1016/j.jocs.2025.102757","url":null,"abstract":"<div><div>With the advantages of real-time computation and transparent decision-making process, Liquid time-constant neural networks (LTCs) perform well in modeling time-varying systems, but they also face problems such as the limitations of dependent learning capabilities and the adaptability to different data distributions. In order to improve the model learning performance, a liquid neural network with physical evolution for variable continuous time series prediction is proposed. Firstly, a simulation synaptic neural transmission model is combined with Hamilton evolution to establish a dynamic evolution model with physical states for neural information transfer. Then an explicit time-dependent Hamiltonian closed-form continuous-depth network (HCFC) is constructed to handle the transmission processing. The implicit Hamilton canonical equations are utilized to model the sophisticated nonlinear transformations experienced by the information as it propagates to the current neuron. The experimental results show that the HCFC model can improve the neural network dynamic property, enhance the learning performance, and provide superior performance in the prediction of complex continuous time series compared with the state-of-the-art (SOTA) methods.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"94 ","pages":"Article 102757"},"PeriodicalIF":3.7,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145659093","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A nonlinear finite volume scheme preserving positivity for the numerical simulation of groundwater contaminant transport in porous media 地下水污染物在多孔介质中运移的非线性有限体积格式

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-10-30 DOI: 10.1016/j.jocs.2025.102739

Sadiq Hamidi , Mustapha El Ossmani , Abdelaziz Taakili

{"title":"A nonlinear finite volume scheme preserving positivity for the numerical simulation of groundwater contaminant transport in porous media","authors":"Sadiq Hamidi , Mustapha El Ossmani , Abdelaziz Taakili","doi":"10.1016/j.jocs.2025.102739","DOIUrl":"10.1016/j.jocs.2025.102739","url":null,"abstract":"<div><div>In this paper, we propose a monotone finite volume scheme to simulate a nonlinear degenerate parabolic problem modeling the transport of reactive solutes in groundwater. Our approach is based on the mixed finite element method (MFEM) for Darcy’s flow equation, combined with a fully implicit Euler nonlinear monotone finite volume scheme for the transport equation. We introduce and analyze the scheme in the context of continuous, non-Lipschitz sorption isotherms, where the problem becomes temporally degenerate, requiring a regularization step to address this issue. The proposed method relies on a Nonlinear Two-Point Flux Approximation (NTPFA) for the diffusive term and an enhanced second-order accurate upwind scheme for the convective term. The fully discrete nonlinear system arising at each time step is solved using Picard iteration, accelerated by the Anderson Acceleration (AA) algorithm. Our aim is, on the one hand, to prove that the scheme preserves the positivity of the analytical solution for anisotropic and heterogeneous full-tensor coefficients on unstructured meshes. On the other hand, we establish the stability, mass conservation, and convergence of the fixed-point algorithm applied to the regularized scheme. To illustrate the effectiveness of our method, several numerical experiments and simulations are presented.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102739"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145420500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Efficient cellular automata-Lattice Boltzmann method for modeling solidification microstructure 高效元胞自动机-晶格玻尔兹曼方法模拟凝固组织

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-10-23 DOI: 10.1016/j.jocs.2025.102735

Jingjing Wang , Xiaoyu Liu , Ruina Mao

{"title":"Efficient cellular automata-Lattice Boltzmann method for modeling solidification microstructure","authors":"Jingjing Wang , Xiaoyu Liu , Ruina Mao","doi":"10.1016/j.jocs.2025.102735","DOIUrl":"10.1016/j.jocs.2025.102735","url":null,"abstract":"<div><div>This paper introduces a 3D GPU-accelerated Cellular Automata-Lattice Boltzmann (CA-LB) Method that resolves the fully coupled interaction between thermosolutal convection and dendritic growth in a single three-dimensional simulation. The novelty of the proposed framework lies in (i) tailoring GPU kernels and memory layouts specifically for the simultaneous evolution of solute, temperature and solid fraction fields, and (ii) revealing three-dimensional helical flow patterns that directly dictate upstream/downstream branch selection and final micro-segregation patterns. Targeted optimizations, lattice-ordered storage of 19 distribution functions and a refactored collision-streaming pattern, yield a 1058 × speed-up over a CPU serial code and enable grid-resolved simulations of Fe–0.6 wt% C alloy solidification within hours instead of weeks. Benchmark validations against natural-convection solutions (error < 0.9 %) and modified Lipton–Glicksman–Kurz analytical predictions confirm quantitative accuracy. The resulting physics highlights that (1) dendrite growth drives convection that preferentially accelerates upstream arms while suppressing downstream arms, (2) growth direction and domain confinement control vortex topology and intensity, and (3) complex helical streamline patterns and dendritic forms can be precisely captured, providing a comprehensive understanding of fluid flow and dendritic growth dynamics, revealing features unseen in 2D simulations. These findings offer an high-throughput platform for designing next-generation alloys with targeted solidification microstructures.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102735"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145371233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A generalized colouring method for a parallelizable integer linear programming approach to polyomino tiling 多线形平铺的可并行整数线性规划的一种广义着色方法

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-10-24 DOI: 10.1016/j.jocs.2025.102734

Marcus R. Garvie , John Burkardt

{"title":"A generalized colouring method for a parallelizable integer linear programming approach to polyomino tiling","authors":"Marcus R. Garvie , John Burkardt","doi":"10.1016/j.jocs.2025.102734","DOIUrl":"10.1016/j.jocs.2025.102734","url":null,"abstract":"<div><div>This article presents a generalized checkerboard colouring method for solving large polyomino tiling problems using integer linear programming (ILP). Extending a previous two-colour approach, our method scales to three or more colours, improving runtimes in cases where the two-colour method offers no advantage. We rigorously derive ILP formulations for both the coloured and uncoloured cases and propose a proof-of-concept parallel implementation to improve scalability in these <span><math><mi>NP</mi></math></span>-complete tiling problems. Numerical experiments using <span>MATLAB</span>, <span>CPLEX</span>, and <span>Gurobi</span> demonstrate significant reductions in problem size and computation time. This approach has the potential to solve tiling problems substantially larger than those previously tractable using ILP-based methods. Although our colouring method generates many subproblems, it allows for the efficient exploration of a manageable subset to find feasible solutions. Our publicly available <span>MATLAB</span> package, <span>CHROMINOES</span> (v1.0.0), constructs ILP formulations, plots solutions, and is available for download on <span><span>Zenodo.org</span><svg><path></path></svg></span>. We conclude by discussing the implications of these results and outlining key challenges in developing a fully practical parallel implementation, including load balancing and managing the overhead associated with processing a large number of subproblems.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102734"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145371234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Thought Management System for long-horizon, goal-driven LLM agents 思想管理系统的长远眼光，目标驱动的法学硕士代理人

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-11-06 DOI: 10.1016/j.jocs.2025.102740

Andrii Bidochko , Yaroslav Vyklyuk

{"title":"Thought Management System for long-horizon, goal-driven LLM agents","authors":"Andrii Bidochko , Yaroslav Vyklyuk","doi":"10.1016/j.jocs.2025.102740","DOIUrl":"10.1016/j.jocs.2025.102740","url":null,"abstract":"<div><div>The “Thoughts Management System” (TMS) is a novel framework that enables autonomous AI Agents to execute long-horizon tasks, inspired by the human mind’s ability to manage 60,000 thoughts per day. TMS empowers agents to dynamically prioritize goals, decompose complex objectives into actionable tasks, and adapt strategies over extended periods. TMS introduces a hierarchical goal decomposition mechanism for breaking down high-level tasks, combined with self-critique modules that iteratively evaluate progress and refine decision-making. By integrating reinforcement learning reward mechanisms, agents focus on high-value tasks while eliminating irrelevant ones, ensuring efficiency and goal alignment.</div><div>Built on our prior work, “LLMAgentNet: A Collaborative Network of Autonomous AI Agents,” TMS uses a multi-agent system (MAS) where specialized agents collaborate to achieve shared goals. Techniques like Monte Carlo Tree Search (MCTS) balance exploration and exploitation for optimal performance in dynamic environments. By mimicking human thought management, TMS enables AI Agents to sustain focus, adapt dynamically, and self-improve over time. This framework marks a significant advancement toward autonomous AI systems capable of solving real-world, long-term challenges across domains such as marketing, scientific discovery, and infrastructure management.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102740"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145579406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Local method of fundamental solutions formulations for polyharmonic BVPs 多谐bvp基本解公式的局部方法

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-11-06 DOI: 10.1016/j.jocs.2025.102742

Andreas Karageorghis , C.S. Chen , Malgorzata A. Jankowska

{"title":"Local method of fundamental solutions formulations for polyharmonic BVPs","authors":"Andreas Karageorghis , C.S. Chen , Malgorzata A. Jankowska","doi":"10.1016/j.jocs.2025.102742","DOIUrl":"10.1016/j.jocs.2025.102742","url":null,"abstract":"<div><div>We develop three local method of fundamental solutions (LMFS) formulations for solving two-dimensional boundary value problems (BVPs) governed by the polyharmonic equation <span><math><mrow><msup><mrow><mi>Δ</mi></mrow><mrow><mi>N</mi></mrow></msup><mi>u</mi><mspace></mspace><mo>=</mo><mspace></mspace><mn>0</mn><mo>,</mo><mspace></mspace><mi>N</mi><mo>∈</mo><mi>N</mi><mo>∖</mo><mrow><mo>{</mo><mn>1</mn><mo>}</mo></mrow></mrow></math></span>, in <span><math><mrow><mi>Ω</mi><mo>⊂</mo><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></mrow></math></span>. The efficacy of the proposed techniques is demonstrated by several numerical experiments in multi-connected domains. Using the local version of the MFS, the difficulty of selecting the source points in multi-connected domains, especially those containing tiny holes, can be alleviated. We also demonstrate the importance of selecting the appropriate formulation for solving higher order polyharmonic equations. In addition, we formulate matrix decomposition algorithms for the efficient solution of polyharmonic BVPs in radially symmetric domains.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102742"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145529199","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhancing urban solar photovoltaic system performance evaluation through a disc spherical fuzzy aggregation framework 通过圆盘球形模糊聚合框架增强城市太阳能光伏系统性能评价

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-11-26 DOI: 10.1016/j.jocs.2025.102758

Shahzaib Ashraf , Muhammad Shakir Chohan , Wania Iqbal , Vladimir Simic , Dragan Pamucar , Nebojsa Bacanin

{"title":"Enhancing urban solar photovoltaic system performance evaluation through a disc spherical fuzzy aggregation framework","authors":"Shahzaib Ashraf , Muhammad Shakir Chohan , Wania Iqbal , Vladimir Simic , Dragan Pamucar , Nebojsa Bacanin","doi":"10.1016/j.jocs.2025.102758","DOIUrl":"10.1016/j.jocs.2025.102758","url":null,"abstract":"<div><div>The integration of solar photovoltaic (PV) systems in urban environments promises great potential for sustainable energy applications. However, the unique characteristics of cities, the varieties of weather that occur at the place, and technology inefficiency make performance evaluation difficult. This paper sought to address the pressing need for a robust performance evaluation framework for urban solar PV systems by developing a disc spherical fuzzy aggregation framework. It develops basic algebraic aggregation operations in the framework of the disc spherical fuzzy set (D-SFSs), proving their completeness and describing their essential characteristics. These new operators conceived to operate on D-SFSs furnish theoretical robustness and provide the foundation for decisions made. A shining novel disc spherical fuzzy method is developed namely combinative distance-based assessment (CODAS) in D-SFS. A case study regarding the application of this model in the assessment of performance by urban solar PV systems is being conducted, thus proving the application aspect. Results come out positive in interpreting the decision-making dilemma and differences among several experts. This would, therefore, encourage various sectors to expand the use of D-SFSs in decision support systems and similar areas by showing how useful they can be in actual situations.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102758"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145623518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Two improved physics-informed Neural Networks for solving Burgers equation 两个改进的物理信息神经网络求解汉堡方程

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-11-24 DOI: 10.1016/j.jocs.2025.102756

Zeyue Zhang , Chunlei Ruan , Zhijun Liu

{"title":"Two improved physics-informed Neural Networks for solving Burgers equation","authors":"Zeyue Zhang , Chunlei Ruan , Zhijun Liu","doi":"10.1016/j.jocs.2025.102756","DOIUrl":"10.1016/j.jocs.2025.102756","url":null,"abstract":"<div><div>Burgers equation, a simplified mathematical model for the Navier–Stokes equation, is often discussed in Computational Fluid Dynamics (CFD). The Burgers equation is a nonlinear convection–diffusion equation in mathematics, its hyperbolic characteristic challenges the numerical methods. Previous studies with the Physics-Informed Neural Networks (PINNs) on Burgers equation have focused on the relatively high viscosity. In that case, the hyperbolicity of Burgers equation is weak and convergent solutions can be obtained. In this study, we attempt to solve Burgers equation with vanishingly small viscosity and with no viscosity. To obtain the convergent and accurate solutions, we present two improved PINNs, one is the PINNs with time segmentation and the other is PINNs with the Weighted Essential Non-oscillatory (WENO) indicator. The former one is to divide the entire time domain into multiple continuous time periods and to iterate training in each time period. The latter is proposed to deal with the discontinuity. The WENO indicator is used to find the non-smooth region in which more training points should be added. Two numerical examples, including one-dimensional Burgers equation with vanishingly small viscosity with Dirichlet boundary conditions and one-dimensional non-viscous Burgers equation with Dirichlet boundary conditions, are carried out. By comparing with the results obtained by the 5th order WENO-Z method and the other three PINNs methods, the validity and high accuracy of the improved PINNs are proved.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102756"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145623520","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Computational model of ammonia synthesis with catalyst pellets: Linking the structure and performance 催化球团合成氨的计算模型：连接结构与性能

IF 3.7 3区计算机科学

Journal of Computational Science Pub Date : 2026-01-01 Epub Date: 2025-10-30 DOI: 10.1016/j.jocs.2025.102737

Michael Abeiku Daniels, Anna Geohagan, Agnieszka Truszkowska

{"title":"Computational model of ammonia synthesis with catalyst pellets: Linking the structure and performance","authors":"Michael Abeiku Daniels, Anna Geohagan, Agnieszka Truszkowska","doi":"10.1016/j.jocs.2025.102737","DOIUrl":"10.1016/j.jocs.2025.102737","url":null,"abstract":"<div><div>Packed bed reactors are one of the most widely utilized reactors in today’s industry, consisting of randomly packed catalyst pellets with reaction participants flowing through the void space and diffusing into the pellets. The efficiency of these reactors is reduced due to a range of factors, and there exist continuous efforts to improve their performance. Due to their complexity, these systems offer numerous opportunities for optimization, relying on computational models to accelerate the exploration of enhancements. The modeling efforts rarely consider the pore space of the reactors, as its accurate representation requires large computational resources. In this work, we leverage an established model of ammonia synthesis to simulate reactors with explicitly resolved catalyst pellets in two dimensions. We reformulate one of the original parameters and identify its suitable value through an optimization algorithm, along with the properties of pellet size distribution that best approximates the actual process. We then use a recently proposed approach for identifying the structural characteristics of porous media that dominate their performance to find which pore-scale properties are likely to have an impact on reactor performance. Once found, we change the suggested properties by manipulating particle radii and their spacing, achieving the expected yield increase. We show that these structural changes can be used to introduce both significant and gradual increases of product yield, which is beneficial for identifying ways to improve the reactor performance and the model itself.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"93 ","pages":"Article 102737"},"PeriodicalIF":3.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145420499","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0