Journal of Engineering Research最新文献_第3页

The role of knowledge engineering in enhancing renewable energy systems for industrial applications 知识工程在加强工业应用的可再生能源系统中的作用

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-10-14 DOI: 10.1016/j.jer.2025.10.006

Suhail H. Serbaya

Transitioning to renewable energy systems (RES) is crucial for reducing industrial carbon emissions and meeting decarbonization targets. While Knowledge Engineering (KE) proves effective in traditional energy systems, its application in optimizing RES in industrial contexts remains underexplored. Existing research lacks comprehensive methodologies for integrating KE with RES to address their inherent variability and unpredictability. This study adopts a quantitative approach, utilizing survey-based data collection and secondary data analysis. KE tools used in industrial RES are classified based on reasoning type, data handling capabilities, and application environments. A sample of at least 100 professionals from the renewable energy sector is surveyed, and statistical techniques, including ANOVA, Chi-square tests, and machine learning models such as decision trees and random forests, are applied to analyze the data. These models are rigorously evaluated using nested cross-validation and multiple performance metrics (accuracy, ROC–AUC, PR–AUC, F1, MCC) to ensure robustness and reliability of findings. When analyzed by reasoning type, Hybrid reasoning tools (n = 18) achieve a mean effectiveness rating of 4.2 in fault detection, significantly outperforming Object-oriented reasoning (n = 15, mean = 3.7) by approximately 15 %. In contrast, pooled results across all KE tools yield lower overall fault detection effectiveness (mean ≈ 2.9–3.0), underscoring the importance of subgroup-specific analysis. Object-oriented reasoning also shows superior performance in energy optimization, achieving a mean rating of 3.8. This research offers novel insights into integrating KE tools to enhance RES in industrial environments. This study provides a novel classification and performance analysis of KE tools. It enhances RES deployment and operational efficiency in industrial settings, directly contributing to sustainable energy access (SDG 7) and climate mitigation (SDG 13).

向可再生能源系统（RES）过渡对于减少工业碳排放和实现脱碳目标至关重要。虽然知识工程（Knowledge Engineering， KE）在传统能源系统中被证明是有效的，但它在工业环境中优化可再生能源方面的应用仍有待探索。现有的研究缺乏综合的方法来整合KE和RES，以解决其固有的可变性和不可预测性。本研究采用定量研究方法，利用调查数据收集和二手数据分析。工业RES中使用的KE工具根据推理类型、数据处理能力和应用程序环境进行分类。调查了来自可再生能源行业的至少100名专业人员的样本，并应用统计技术，包括方差分析，卡方检验以及决策树和随机森林等机器学习模型来分析数据。这些模型使用嵌套交叉验证和多个性能指标（准确性、ROC-AUC、PR-AUC、F1、MCC）进行严格评估，以确保结果的稳健性和可靠性。当按推理类型进行分析时，混合推理工具（n = 18）在故障检测方面的平均有效性评分为4.2，显著优于面向对象推理（n = 15，平均值= 3.7）约15 %。相比之下，所有KE工具的汇总结果产生较低的总体故障检测效率（平均值≈2.9-3.0），强调了亚组特定分析的重要性。面向对象推理在能源优化方面也显示出卓越的性能，达到了3.8的平均评分。本研究为集成KE工具以增强工业环境中的RES提供了新颖的见解。本研究提供了一种新的KE工具分类和性能分析方法。它提高了可再生能源在工业环境中的部署和运营效率，直接促进了可持续能源获取（可持续发展目标7）和减缓气候变化（可持续发展目标13）。

{"title":"The role of knowledge engineering in enhancing renewable energy systems for industrial applications","authors":"Suhail H. Serbaya","doi":"10.1016/j.jer.2025.10.006","DOIUrl":"10.1016/j.jer.2025.10.006","url":null,"abstract":"<div><div>Transitioning to renewable energy systems (RES) is crucial for reducing industrial carbon emissions and meeting decarbonization targets. While Knowledge Engineering (KE) proves effective in traditional energy systems, its application in optimizing RES in industrial contexts remains underexplored. Existing research lacks comprehensive methodologies for integrating KE with RES to address their inherent variability and unpredictability. This study adopts a quantitative approach, utilizing survey-based data collection and secondary data analysis. KE tools used in industrial RES are classified based on reasoning type, data handling capabilities, and application environments. A sample of at least 100 professionals from the renewable energy sector is surveyed, and statistical techniques, including ANOVA, Chi-square tests, and machine learning models such as decision trees and random forests, are applied to analyze the data. These models are rigorously evaluated using nested cross-validation and multiple performance metrics (accuracy, ROC–AUC, PR–AUC, F1, MCC) to ensure robustness and reliability of findings. When analyzed by reasoning type, Hybrid reasoning tools (n = 18) achieve a mean effectiveness rating of 4.2 in fault detection, significantly outperforming Object-oriented reasoning (n = 15, mean = 3.7) by approximately 15 %. In contrast, pooled results across all KE tools yield lower overall fault detection effectiveness (mean ≈ 2.9–3.0), underscoring the importance of subgroup-specific analysis. Object-oriented reasoning also shows superior performance in energy optimization, achieving a mean rating of 3.8. This research offers novel insights into integrating KE tools to enhance RES in industrial environments. This study provides a novel classification and performance analysis of KE tools. It enhances RES deployment and operational efficiency in industrial settings, directly contributing to sustainable energy access (SDG 7) and climate mitigation (SDG 13).</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 1140-1152"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453901","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A stability prediction model for entry-type excavations based on Adaptive Boosting and swarm intelligence optimization 基于自适应助推和群智能优化的入口式开挖稳定性预测模型

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-08-05 DOI: 10.1016/j.jer.2025.08.001

Mudan Wu

With increasing excavation depths, geological heterogeneity intensifies, rendering the reliable prediction of entry-type excavation stability an urgent yet formidable challenge in underground engineering. In this study, an innovative stability prediction framework was proposed for entry-type excavations that couples Adaptive Boosting (AdaBoost) with the Whale Optimization Algorithm (WOA), Pelican Optimization Algorithm (POA) and Zebra Optimization Algorithm (ZOA). Each algorithm sustains an effective exploration-exploitation balance, as evidenced by convergent fitness and diversity trajectories that preclude premature stagnation. Rigorous validation demonstrates that all hybrid models significantly outperform conventional classifiers, with POA-AdaBoost and ZOA-AdaBoost achieving the highest scores (Accuracy: 0.883, Precision: 0.895, Recall: 0.830, F1 score: 0.855, Kappa coefficient: 0.781). Balancing predictive fidelity with computational efficiency, the POA-AdaBoost model (population size = 100, n_estimators = 74, learning_rate = 1.220) is identified as optimal. Under complex geological and operational scenarios, this model further delivers more nuanced stability delineation than traditional empirical criteria, surpassing Kumar’s method in classifying unstable cases within high-risk domains (alongside a 43.75 % reduction in unsafe misclassifications). These findings highlight the promise of integrating swarm intelligence with ensemble learning to advance intelligent stability assessment in underground excavations.

随着开挖深度的增加，地质非均质性加剧，使入口式开挖稳定性的可靠预测成为地下工程中迫切而艰巨的挑战。本研究提出了一种将自适应推进（AdaBoost）与鲸鱼优化算法（WOA）、鹈鹕优化算法（POA）和斑马优化算法（ZOA）相结合的创新型入口型基坑稳定性预测框架。每个算法都保持有效的探索-开发平衡，收敛适应度和多样性轨迹可以防止过早停滞。严格的验证表明，所有混合模型都显著优于传统分类器，其中POA-AdaBoost和ZOA-AdaBoost获得了最高分（准确率：0.883，精度：0.895，召回率：0.830，F1分数：0.855，Kappa系数：0.781）。通过权衡预测保真度和计算效率，POA-AdaBoost模型（population size = 100, n_estimators = 74, learning_rate = 1.220）被确定为最优模型。在复杂的地质和操作场景下，该模型进一步提供了比传统经验标准更细致的稳定性描述，在高风险域内对不稳定情况进行分类方面超过了Kumar的方法（同时不安全错误分类减少43.75% %）。这些发现突出了将群体智能与集成学习相结合来推进地下开挖稳定性智能评估的前景。

{"title":"A stability prediction model for entry-type excavations based on Adaptive Boosting and swarm intelligence optimization","authors":"Mudan Wu","doi":"10.1016/j.jer.2025.08.001","DOIUrl":"10.1016/j.jer.2025.08.001","url":null,"abstract":"<div><div>With increasing excavation depths, geological heterogeneity intensifies, rendering the reliable prediction of entry-type excavation stability an urgent yet formidable challenge in underground engineering. In this study, an innovative stability prediction framework was proposed for entry-type excavations that couples Adaptive Boosting (AdaBoost) with the Whale Optimization Algorithm (WOA), Pelican Optimization Algorithm (POA) and Zebra Optimization Algorithm (ZOA). Each algorithm sustains an effective exploration-exploitation balance, as evidenced by convergent fitness and diversity trajectories that preclude premature stagnation. Rigorous validation demonstrates that all hybrid models significantly outperform conventional classifiers, with POA-AdaBoost and ZOA-AdaBoost achieving the highest scores (Accuracy: 0.883, Precision: 0.895, Recall: 0.830, F1 score: 0.855, Kappa coefficient: 0.781). Balancing predictive fidelity with computational efficiency, the POA-AdaBoost model (population size = 100, n_estimators = 74, learning_rate = 1.220) is identified as optimal. Under complex geological and operational scenarios, this model further delivers more nuanced stability delineation than traditional empirical criteria, surpassing Kumar’s method in classifying unstable cases within high-risk domains (alongside a 43.75 % reduction in unsafe misclassifications). These findings highlight the promise of integrating swarm intelligence with ensemble learning to advance intelligent stability assessment in underground excavations.</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 358-369"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453904","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

3D Double diffusive natural convection in an open pyramidal cavity: CFD study and machine learning prediction 开放锥体腔内三维双扩散自然对流：CFD研究和机器学习预测

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-10-17 DOI: 10.1016/j.jer.2025.10.009

Inès Hilali-Jaghdam , Walid Aich , Amnah Alshahrani , Chemseddine Maatki , Badr M. Alshammari , Lioua Kolsi

The present study corresponds to a CFD based analysis of the three-dimensional double-diffusive natural convection in an open pyramidal cavity. The equations governing the fluid flow, heat and mass transfer and entropy generation within the cavity are solved using the Finite Element Method. In addition, machine learning (ML) models were developed to estimate the average Nusselt number Nu_av, the average Sherwood number Sh_av and the total generated entropy S_tot as functions of governing parameters in order to enable fast and reliable prediction of the system's thermal and mass transport performance. Thus, the Random Forest regression model is trained on the CFD results as a surrogate predictor. The accuracy of the predictive model is excellent, ensuring R² > 0.96 for all output targets, and the main findings indicate that the optimal configuration occurs at high buoyancy ratio (N > 1) and wide cavity openings, which leads to enhanced heat and mass transfer. The minimal entropy generation occurs at N = –1, where the opposing thermal and solutal forces reduces the irreversibilities. Overall, it can be mentioned that the integration of machine-learning approach with CFD leads to robust predictive models allowing fast and accurate design of the double-diffusive convection in 3D open geometries. Increasing the cavity opening width L significantly enhances heat and mass transfer: for example, in a strongly opposing-buoyancy case (N = –2) Nu_av more than doubles (from 3.8 to 8.1) and Sh_av nearly quintuples (from 1.7 to 8.7) as L increases from 0.1 to 1.0, with similarly high absolute values achieved under aiding buoyancy (e.g., Nu_av ≈8.5 and Sh_av ≈12.1 at L = 1.0 for N = +2).

本研究对应于基于CFD的开放锥体腔内三维双扩散自然对流的分析。采用有限元法求解了腔内流体流动、传热传质和熵生成方程。此外，开发了机器学习（ML）模型来估计平均努塞尔数Nuav，平均舍伍德数Shav和总生成熵Stot作为控制参数的函数，以便能够快速可靠地预测系统的热传输和质量传输性能。因此，随机森林回归模型在CFD结果上作为替代预测器进行训练。预测模型的精度非常好，确保了所有输出目标的R2 >； 0.96，主要研究结果表明，在高浮力比（N >； 1）和宽空腔开度时，最优配置会导致传热传质增强。最小熵产生发生在N = -1，在那里相反的热力和溶质力减少了不可逆性。总的来说，可以提到的是，机器学习方法与CFD的集成导致了强大的预测模型，从而可以快速准确地设计3D开放几何中的双扩散对流。增加空腔宽度L显著增强传热传质：例如，在强反浮力情况下（N = -2），当L从0.1增加到1.0时，Nuav增加了一倍多（从3.8到8.1），Shav增加了近五倍（从1.7到8.7），在辅助浮力下也获得了同样高的绝对值（例如，当N = +2时，L = 1.0时，Nuav≈8.5,Shav≈12.1）。

{"title":"3D Double diffusive natural convection in an open pyramidal cavity: CFD study and machine learning prediction","authors":"Inès Hilali-Jaghdam , Walid Aich , Amnah Alshahrani , Chemseddine Maatki , Badr M. Alshammari , Lioua Kolsi","doi":"10.1016/j.jer.2025.10.009","DOIUrl":"10.1016/j.jer.2025.10.009","url":null,"abstract":"<div><div>The present study corresponds to a CFD based analysis of the three-dimensional double-diffusive natural convection in an open pyramidal cavity. The equations governing the fluid flow, heat and mass transfer and entropy generation within the cavity are solved using the Finite Element Method. In addition, machine learning (ML) models were developed to estimate the average Nusselt number Nu<sub>av</sub>, the average Sherwood number Sh<sub>av</sub> and the total generated entropy S<sub>tot</sub> as functions of governing parameters in order to enable fast and reliable prediction of the system's thermal and mass transport performance. Thus, the Random Forest regression model is trained on the CFD results as a surrogate predictor. The accuracy of the predictive model is excellent, ensuring R<sup>2</sup> > 0.96 for all output targets, and the main findings indicate that the optimal configuration occurs at high buoyancy ratio (N > 1) and wide cavity openings, which leads to enhanced heat and mass transfer. The minimal entropy generation occurs at N = –1, where the opposing thermal and solutal forces reduces the irreversibilities. Overall, it can be mentioned that the integration of machine-learning approach with CFD leads to robust predictive models allowing fast and accurate design of the double-diffusive convection in 3D open geometries. Increasing the cavity opening width L significantly enhances heat and mass transfer: for example, in a strongly opposing-buoyancy case (N = –2) Nu<sub>av</sub> more than doubles (from 3.8 to 8.1) and Sh<sub>av</sub> nearly quintuples (from 1.7 to 8.7) as L increases from 0.1 to 1.0, with similarly high absolute values achieved under aiding buoyancy (e.g., Nu<sub>av</sub> ≈8.5 and Sh<sub>av</sub> ≈12.1 at L = 1.0 for N = +2).</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 577-594"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhanced intrusion detection system using feature selection and hybrid learning models for high performance and efficiency in an IOT environment 使用特征选择和混合学习模型的增强入侵检测系统，在物联网环境中实现高性能和高效率

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-10-26 DOI: 10.1016/j.jer.2025.10.016

Omar Achbarou , Toufik Datsi , Outmane Bourkoukou , Ahmed My El kiram

Attacks on automated systems are increasingly frequent, sophisticated, intelligent and distributed, posing serious threats to an Internet of Things (IoT) environment and necessitating costly interventions. Intrusion detection systems (IDSs) therefore play an essential role in protecting IoT environments against these types of attack. However, traditional IDS detection methods, which are mainly based on signatures and rules, face significant challenges such as an inability to detect unknown attacks, resulting in low detection rates, high false positive rates and long response times.

To overcome these limitations, this paper proposes a novel deep learning approach for binary classification of network traffic (normal vs. attack) in the context of anomaly detection, using convolutional neural networks (CNNs) and feature selection techniques for an IDS. A CNN is leveraged for its ability to detect spatial patterns in network characteristics, a crucial capability in dynamic and distributed IoT environments where the modeling of normal behavior is challenging. To mitigate data imbalance, our model applies a two-stage strategy: first, the Synthetic Minority Oversampling Technique (SMOTE) is used for global balancing before dataset splitting, and second, a class weighting technique is applied after splitting, during training. The aim of this approach is to enhance detection accuracy and reduce the execution time, to provide a more effective IDS for IoT networks.

We conduct experiments on CICIDS2017, a widely used benchmark dataset, to evaluate and compare the performance of Random Forest (RF), LightGBM and XGBoost classifiers, both with and without feature selection. The results show remarkable improvements in regard to accuracy (99.95 %), precision (99.99 %), and recall (99.91 %), especially with the XGBoost model with 40 selected features, while maintaining a very competitive runtime that is significantly lower than comparable methods in existing studies. Furthermore, our model outperforms other existing schemes in terms of performance, efficiency and scalability.

针对自动化系统的攻击越来越频繁、复杂、智能和分布式，对物联网（IoT）环境构成严重威胁，需要昂贵的干预措施。因此，入侵检测系统（ids）在保护物联网环境免受这些类型的攻击方面发挥着至关重要的作用。传统的IDS检测方法主要基于签名和规则，无法检测到未知攻击，导致检测率低、误报率高、响应时间长。为了克服这些限制，本文提出了一种新的深度学习方法，用于异常检测背景下的网络流量（正常与攻击）的二进制分类，使用卷积神经网络（cnn）和IDS的特征选择技术。CNN具有检测网络特征空间模式的能力，这在动态和分布式物联网环境中是一项至关重要的能力，在这些环境中，正常行为的建模具有挑战性。为了减轻数据不平衡，我们的模型采用了两阶段策略：首先，在数据集分裂之前使用合成少数过采样技术（SMOTE）进行全局平衡，其次，在分裂之后，在训练期间应用类加权技术。这种方法的目的是提高检测精度，减少执行时间，为物联网网络提供更有效的IDS。我们在广泛使用的基准数据集CICIDS2017上进行了实验，以评估和比较随机森林（RF）、LightGBM和XGBoost分类器在有和没有特征选择的情况下的性能。结果显示，在准确率（99.95 %）、精密度（99.99 %）和召回率（99.91 %）方面有了显著的提高，特别是在具有40个选定特征的XGBoost模型上，同时保持了一个非常有竞争力的运行时间，显著低于现有研究中的同类方法。此外，我们的模型在性能、效率和可扩展性方面优于其他现有方案。

{"title":"Enhanced intrusion detection system using feature selection and hybrid learning models for high performance and efficiency in an IOT environment","authors":"Omar Achbarou , Toufik Datsi , Outmane Bourkoukou , Ahmed My El kiram","doi":"10.1016/j.jer.2025.10.016","DOIUrl":"10.1016/j.jer.2025.10.016","url":null,"abstract":"<div><div>Attacks on automated systems are increasingly frequent, sophisticated, intelligent and distributed, posing serious threats to an Internet of Things (IoT) environment and necessitating costly interventions. Intrusion detection systems (IDSs) therefore play an essential role in protecting IoT environments against these types of attack. However, traditional IDS detection methods, which are mainly based on signatures and rules, face significant challenges such as an inability to detect unknown attacks, resulting in low detection rates, high false positive rates and long response times.</div><div>To overcome these limitations, this paper proposes a novel deep learning approach for binary classification of network traffic (normal vs. attack) in the context of anomaly detection, using convolutional neural networks (CNNs) and feature selection techniques for an IDS. A CNN is leveraged for its ability to detect spatial patterns in network characteristics, a crucial capability in dynamic and distributed IoT environments where the modeling of normal behavior is challenging. To mitigate data imbalance, our model applies a two-stage strategy: first, the Synthetic Minority Oversampling Technique (SMOTE) is used for global balancing before dataset splitting, and second, a class weighting technique is applied after splitting, during training. The aim of this approach is to enhance detection accuracy and reduce the execution time, to provide a more effective IDS for IoT networks.</div><div>We conduct experiments on CICIDS2017, a widely used benchmark dataset, to evaluate and compare the performance of Random Forest (RF), LightGBM and XGBoost classifiers, both with and without feature selection. The results show remarkable improvements in regard to accuracy (99.95 %), precision (99.99 %), and recall (99.91 %), especially with the XGBoost model with 40 selected features, while maintaining a very competitive runtime that is significantly lower than comparable methods in existing studies. Furthermore, our model outperforms other existing schemes in terms of performance, efficiency and scalability.</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 953-965"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Evaluation of machine learning models for streamflow projections under different greenhouse gas emission scenarios 不同温室气体排放情景下流量预测的机器学习模型评估

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-06-24 DOI: 10.1016/j.jer.2025.06.006

Abdulhadi Pala, Seydanur Sebcioglu Mutlu, Aytac Guven

With the increasing demand for precise water management under climate change, hydrological prediction has become more important than ever. In this context, machine learning emerges as a powerful tool to improve forecasting performance. This study evaluates the performance of LightGBM, Random Forest (RF) and Ridge Regression (RR) models for streamflow prediction and climate change impact assessment. Using historical streamflow data from 1984 to 2014, model accuracy was assessed with key performance metrics. LightGBM outperformed RF and RR, achieving lower RMSE (3.046 m³/s), MAE (2.145 m³/s), RSR (0.599), and higher NSE (0.64) and KGE (0.683), demonstrating greater accuracy and robustness due to its gradient-boosting framework. Based on these results, LightGBM was used to project future streamflow under CMIP6 climate scenarios with the NorESM2 model. Projections show increasing streamflow trends under SSP3–7.0 (3.436 %) and SSP5–8.5 (4.049 %), a decrease under SSP1–2.6, and a stable trend under SSP2–4.5, according to historical streamflow data. The projected future streamflow under these climate scenarios provides valuable insights for water resource management, highlighting the importance of using advanced machine learning techniques like LightGBM for hydrological forecasting in the context of climate change.

随着气候变化对水资源精准管理的要求越来越高，水文预测变得越来越重要。在这种情况下，机器学习成为提高预测性能的有力工具。本研究评估了LightGBM、Random Forest （RF）和Ridge Regression （RR）模型在河流流量预测和气候变化影响评估中的性能。利用1984 - 2014年的历史流量数据，用关键性能指标评估模型的准确性。LightGBM优于RF和RR，实现了较低的RMSE(3.046 m³/s)， MAE(2.145 m³/s)， RSR(0.599)，较高的NSE（0.64）和KGE(0.683)，由于其梯度增强框架，显示出更高的准确性和鲁棒性。在此基础上，利用LightGBM与NorESM2模式对CMIP6气候情景下的未来流量进行了预估。预测结果显示，在SSP3-7.0（3.436 %）和SSP5-8.5（4.049 %）条件下，流量呈增加趋势，在SSP1-2.6条件下呈减少趋势，在SSP2-4.5条件下呈稳定趋势。在这些气候情景下预测的未来流量为水资源管理提供了有价值的见解，强调了在气候变化背景下使用先进的机器学习技术（如LightGBM）进行水文预测的重要性。

{"title":"Evaluation of machine learning models for streamflow projections under different greenhouse gas emission scenarios","authors":"Abdulhadi Pala, Seydanur Sebcioglu Mutlu, Aytac Guven","doi":"10.1016/j.jer.2025.06.006","DOIUrl":"10.1016/j.jer.2025.06.006","url":null,"abstract":"<div><div>With the increasing demand for precise water management under climate change, hydrological prediction has become more important than ever. In this context, machine learning emerges as a powerful tool to improve forecasting performance. This study evaluates the performance of LightGBM, Random Forest (RF) and Ridge Regression (RR) models for streamflow prediction and climate change impact assessment. Using historical streamflow data from 1984 to 2014, model accuracy was assessed with key performance metrics. LightGBM outperformed RF and RR, achieving lower RMSE (3.046 m³/s), MAE (2.145 m³/s), RSR (0.599), and higher NSE (0.64) and KGE (0.683), demonstrating greater accuracy and robustness due to its gradient-boosting framework. Based on these results, LightGBM was used to project future streamflow under CMIP6 climate scenarios with the NorESM2 model. Projections show increasing streamflow trends under SSP3–7.0 (3.436 %) and SSP5–8.5 (4.049 %), a decrease under SSP1–2.6, and a stable trend under SSP2–4.5, according to historical streamflow data. The projected future streamflow under these climate scenarios provides valuable insights for water resource management, highlighting the importance of using advanced machine learning techniques like LightGBM for hydrological forecasting in the context of climate change.</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 54-68"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147454114","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhancing downlink NOMA performance: A pareto optimization approach for superior power allocation 提高下行NOMA性能：一种帕累托优化方法，用于优越的功率分配

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-08-29 DOI: 10.1016/j.jer.2025.08.021

G. Shanmugavel, M.S. Vasanthi

This paper proposes a novel Pareto-Optimization Algorithm (POA) for dynamic power allocation in downlink Non-Orthogonal Multiple Access (NOMA) systems operating within heterogeneous network environments. The proposed POA framework effectively addresses the critical trade-off between spectral efficiency and user fairness by integrating adaptive modulation schemes (BPSK, QPSK, 16-QAM, 64-QAM) with demand-aware resource categorisation based on Signal-to-Noise-plus-Interference Ratios (SNIR) and User Equipment (UE) requirements. The optimisation process utilises Lagrange multipliers and Karush-Kuhn-Tucker (KKT) conditions to ensure fair and optimal power allocation under total power constraints. Extensive simulation results demonstrate that the proposed POA achieves substantial performance improvements, attaining a 92 % fairness index (versus 71 % for Water Filling), a spectral efficiency of 5.3 bps/Hz, and a peak data rate of 390 Mbps. A comparative analysis against existing schemes, including Fixed Power Allocation (FPA), Fractional Transmit Power Allocation (FTPA), Water Filling Algorithm (WFA), and Particle Swarm Optimisation Algorithm (PSOA), confirms the superiority of POA in enhancing fairness-efficiency trade-offs. These results highlight the POA's suitability for dynamic, heterogeneous 5 G and beyond wireless networks, offering improved scalability and real-time adaptability under diverse channel and traffic conditions.

针对异构网络环境下下行非正交多址（NOMA）系统的动态功率分配问题，提出了一种新的pareto优化算法。提出的POA框架通过将自适应调制方案（BPSK、QPSK、16-QAM、64-QAM）与基于信噪比（SNIR）和用户设备（UE）要求的需求感知资源分类相结合，有效地解决了频谱效率和用户公平性之间的关键权衡。优化过程利用拉格朗日乘数和Karush-Kuhn-Tucker （KKT）条件来确保在总功率约束下公平和最佳的功率分配。大量的仿真结果表明，所提出的POA实现了实质性的性能改进，达到了92% %的公平指数（与71 %的水填充相比），频谱效率为5.3 bps/Hz，峰值数据速率为390 Mbps。通过与现有的固定功率分配（FPA）、分数传输功率分配（FTPA）、充水算法（WFA）和粒子群优化算法（PSOA）的比较分析，证实了POA在提高公平效率权衡方面的优势。这些结果突出了POA对动态、异构5 G及以上无线网络的适用性，在不同信道和流量条件下提供了改进的可扩展性和实时适应性。

{"title":"Enhancing downlink NOMA performance: A pareto optimization approach for superior power allocation","authors":"G. Shanmugavel, M.S. Vasanthi","doi":"10.1016/j.jer.2025.08.021","DOIUrl":"10.1016/j.jer.2025.08.021","url":null,"abstract":"<div><div>This paper proposes a novel Pareto-Optimization Algorithm (POA) for dynamic power allocation in downlink Non-Orthogonal Multiple Access (NOMA) systems operating within heterogeneous network environments. The proposed POA framework effectively addresses the critical trade-off between spectral efficiency and user fairness by integrating adaptive modulation schemes (BPSK, QPSK, 16-QAM, 64-QAM) with demand-aware resource categorisation based on Signal-to-Noise-plus-Interference Ratios (SNIR) and User Equipment (UE) requirements. The optimisation process utilises Lagrange multipliers and Karush-Kuhn-Tucker (KKT) conditions to ensure fair and optimal power allocation under total power constraints. Extensive simulation results demonstrate that the proposed POA achieves substantial performance improvements, attaining a 92 % fairness index (versus 71 % for Water Filling), a spectral efficiency of 5.3 bps/Hz, and a peak data rate of 390 Mbps. A comparative analysis against existing schemes, including Fixed Power Allocation (FPA), Fractional Transmit Power Allocation (FTPA), Water Filling Algorithm (WFA), and Particle Swarm Optimisation Algorithm (PSOA), confirms the superiority of POA in enhancing fairness-efficiency trade-offs. These results highlight the POA's suitability for dynamic, heterogeneous 5 G and beyond wireless networks, offering improved scalability and real-time adaptability under diverse channel and traffic conditions.</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 858-869"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147454244","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Streamlining nitrogen removal in Kuwait’s WWTP: A data-driven analysis of BNR process optimization 简化科威特污水处理厂的脱氮：BNR工艺优化的数据驱动分析

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-10-01 DOI: 10.1016/j.jer.2025.09.017

Noor Alkandari , Nayef Al-Mutairi , Mohammed Hamoda

The elimination of nutrients, especially nitrogen, is an increasingly pressing concern for wastewater treatment facilities (WWTPs). The Sulaibiya WWTP in Kuwait utilizes the conventional Biological Nitrogen Removal (BNR) technique of nitrification-denitrification for secondary treatment. Both Nitrogenous Oxygen Demand (NOD) and Chemical Oxygen Demand (COD) are successfully reduced by this energy-intensive method. The facility employs Reverse Osmosis (RO) for advanced treatment, which, although effective in eliminating all impurities, incurs significant operating expenses, increases energy consumption, and generates brine as a byproduct. This study aims to enhance nitrogen removal efficiency through four optimization scenarios, focusing on minimizing operational costs, reducing energy consumption, eliminating the necessity for tertiary treatment, and ensuring adherence to Kuwait Environment Public Authority (KEPA) standards for irrigation and marine disposal. To assess the possibility of exceeding KEPA standards across five treatment scenarios—the basic model, A²/O, A/O, Modified Ludzack Ettinger (MLE), and 4-stage Bardenpho—the study comprised six years of data analysis, modeling of wastewater treatment facilities using GPS-X software, sensitivity analysis, and Monte Carlo simulation. The existing design of the wastewater treatment facility consistently complies with KEPA standards, attaining removal efficiencies of 92.7 % for TKN, 95.8 % for total suspended solids (TSS), 94.3 % for COD, and 95.3 % for ammonia nitrogen, even in the absence of tertiary treatment. The research demonstrates that the elimination of the tertiary treatment phase is achievable. The facility's optimal configurations were the A/O process, with removal efficiencies of 97.7 % for TKN, 98.3 % for TSS, 96.4 % for COD, and 98.4 % for ammonia, and the MLE process, with efficiencies of 97.7 %, 96.8 %, 96.4 %, and 98.4 %, respectively.

消除营养物，特别是氮，是废水处理设施（WWTPs）日益迫切关注的问题。科威特苏莱比亚污水处理厂采用传统的硝化-反硝化生物脱氮（BNR）技术进行二次处理。通过这种高能耗的方法，氮需氧量（NOD）和化学需氧量（COD）都得到了成功的降低。该设施采用反渗透（RO）进行高级处理，虽然可以有效地消除所有杂质，但会产生大量的运营费用，增加能源消耗，并产生卤水作为副产品。该研究旨在通过四种优化方案提高氮去除效率，重点是最大限度地降低运营成本，降低能耗，消除三级处理的必要性，并确保符合科威特环境公共管理局（KEPA）的灌溉和海洋处理标准。为了评估五个处理方案（基本模型、A2/O、A/O、修正Ludzack Ettinger （MLE）和四阶段bardenph）超过KEPA标准的可能性，该研究包括六年的数据分析、使用GPS-X软件对废水处理设施进行建模、敏感性分析和蒙特卡罗模拟。污水处理设施的现有设计始终符合KEPA标准，即使在没有三级处理的情况下，TKN的去除率为92.7 %，总悬浮固体（TSS）的去除率为95.8% %，COD的去除率为94.3 %，氨氮的去除率为95.3 %。研究表明，消除三级处理阶段是可以实现的。最优配置为A/O工艺对TKN、TSS、COD和氨的去除率分别为97.7% %、98.3% %、96.4 %和98.4% %；MLE工艺对TKN、TSS、COD和氨的去除率分别为97.7% %、96.8% %、96.4 %和98.4% %。

{"title":"Streamlining nitrogen removal in Kuwait’s WWTP: A data-driven analysis of BNR process optimization","authors":"Noor Alkandari , Nayef Al-Mutairi , Mohammed Hamoda","doi":"10.1016/j.jer.2025.09.017","DOIUrl":"10.1016/j.jer.2025.09.017","url":null,"abstract":"<div><div>The elimination of nutrients, especially nitrogen, is an increasingly pressing concern for wastewater treatment facilities (WWTPs). The Sulaibiya WWTP in Kuwait utilizes the conventional Biological Nitrogen Removal (BNR) technique of nitrification-denitrification for secondary treatment. Both Nitrogenous Oxygen Demand (NOD) and Chemical Oxygen Demand (COD) are successfully reduced by this energy-intensive method. The facility employs Reverse Osmosis (RO) for advanced treatment, which, although effective in eliminating all impurities, incurs significant operating expenses, increases energy consumption, and generates brine as a byproduct. This study aims to enhance nitrogen removal efficiency through four optimization scenarios, focusing on minimizing operational costs, reducing energy consumption, eliminating the necessity for tertiary treatment, and ensuring adherence to Kuwait Environment Public Authority (KEPA) standards for irrigation and marine disposal. To assess the possibility of exceeding KEPA standards across five treatment scenarios—the basic model, A<sup>2</sup>/O, A/O, Modified Ludzack Ettinger (MLE), and 4-stage Bardenpho—the study comprised six years of data analysis, modeling of wastewater treatment facilities using GPS-X software, sensitivity analysis, and Monte Carlo simulation. The existing design of the wastewater treatment facility consistently complies with KEPA standards, attaining removal efficiencies of 92.7 % for TKN, 95.8 % for total suspended solids (TSS), 94.3 % for COD, and 95.3 % for ammonia nitrogen, even in the absence of tertiary treatment. The research demonstrates that the elimination of the tertiary treatment phase is achievable. The facility's optimal configurations were the A/O process, with removal efficiencies of 97.7 % for TKN, 98.3 % for TSS, 96.4 % for COD, and 98.4 % for ammonia, and the MLE process, with efficiencies of 97.7 %, 96.8 %, 96.4 %, and 98.4 %, respectively.</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 176-188"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453728","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Decoding the dynamics of daily total global solar radiation: Computational learning strategies powered by explainable AI 解码每日全球太阳总辐射的动态：由可解释的人工智能驱动的计算学习策略

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-10-15 DOI: 10.1016/j.jer.2025.10.008

Muhammed Ernur Akiner , Naim Süleyman Tınğ , Abdullah A. Alsumaiei , Okan Mert Katipoğlu

This study investigates the prediction of daily total global solar radiation using advanced machine learning and hybrid models. The input variables include various meteorological and environmental parameters such as total precipitation, average snow height, total solar intensity, existing snow height, soil temperature, actual pressure, average temperature, cloudiness, minimum cloud height, and relative humidity. Based on SHAP value analysis, 5 cm Soil Temperature (+67.08) and Total Solar Intensity (+34.76) emerged as the most influential features for the models, while parameters like Current Snow Depth (+0.33) and Average Snow Depth (+0.36) contributed minimally. The analysis spans a period from 2006 to 2022 using daily data. The dataset is divided into 70 % for training and 20 % for validation, with 10 % of the training data allocated for testing. Several models were employed, including AdaBoost, CatBoost, Deep Belief Network (DBN), Light Gradient Boosting Machine (LGBM), Adaptive Neuro-Fuzzy Inference System (ANFIS), Echo State Neural Networks (ESN), and Autoencoder-based regression. According to the analysis results, the Catboost model produces the most accurate predictions with RMSE: 97.25 KW.hr/m², MAE: 73.64 KW.hr/m², AIC: 17075, NSE: 0.65 and, KGE: 0.72 values, while the DBN model shows the second best model with RMSE: 98 KW.hr/m², MAE: 74.48 KW.hr/m², AIC: 17103, NSE: 0.64 and, KGE: 0.74. The analysis results have the potential to provide information to decision makers and planners, especially in the evaluation of the performance of solar panels, the optimization of agricultural processes, and the development of climate change adaptation strategies.

本研究利用先进的机器学习和混合模型研究了每日全球太阳总辐射的预测。输入变量包括各种气象和环境参数，如总降水量、平均雪高、太阳总强度、现有雪高、土壤温度、实际压力、平均温度、云量、最低云高和相对湿度。基于SHAP值分析，5 cm土壤温度（+67.08）和太阳总强度（+34.76）对模型的影响最大，而当前雪深（+0.33）和平均雪深（+0.36）对模型的影响最小。该分析使用每日数据，时间跨度从2006年到2022年。数据集分为70 %用于训练和20 %用于验证，其中10 %的训练数据分配用于测试。采用了AdaBoost、CatBoost、深度信念网络（DBN）、光梯度增强机（LGBM）、自适应神经模糊推理系统（ANFIS）、回声状态神经网络（ESN）和基于自编码器的回归等模型。结果表明，Catboost模型的预测结果最准确，RMSE为97.25 KW.hr/m2， MAE为73.64 KW.hr/m2， AIC为17075，NSE为0.65，KGE为0.72；DBN模型的预测结果次之，RMSE为98 KW.hr/m2， MAE为74.48 KW.hr/m2， AIC为17103，NSE为0.64，KGE为0.74。分析结果有可能为决策者和规划者提供信息，特别是在评估太阳能电池板的性能、优化农业过程和制定气候变化适应战略方面。

{"title":"Decoding the dynamics of daily total global solar radiation: Computational learning strategies powered by explainable AI","authors":"Muhammed Ernur Akiner , Naim Süleyman Tınğ , Abdullah A. Alsumaiei , Okan Mert Katipoğlu","doi":"10.1016/j.jer.2025.10.008","DOIUrl":"10.1016/j.jer.2025.10.008","url":null,"abstract":"<div><div>This study investigates the prediction of daily total global solar radiation using advanced machine learning and hybrid models. The input variables include various meteorological and environmental parameters such as total precipitation, average snow height, total solar intensity, existing snow height, soil temperature, actual pressure, average temperature, cloudiness, minimum cloud height, and relative humidity. Based on SHAP value analysis, 5 cm Soil Temperature (+67.08) and Total Solar Intensity (+34.76) emerged as the most influential features for the models, while parameters like Current Snow Depth (+0.33) and Average Snow Depth (+0.36) contributed minimally. The analysis spans a period from 2006 to 2022 using daily data. The dataset is divided into 70 % for training and 20 % for validation, with 10 % of the training data allocated for testing. Several models were employed, including AdaBoost, CatBoost, Deep Belief Network (DBN), Light Gradient Boosting Machine (LGBM), Adaptive Neuro-Fuzzy Inference System (ANFIS), Echo State Neural Networks (ESN), and Autoencoder-based regression. According to the analysis results, the Catboost model produces the most accurate predictions with RMSE: 97.25 KW.hr/m<sup>2</sup>, MAE: 73.64 KW.hr/m<sup>2</sup>, AIC: 17075, NSE: 0.65 and, KGE: 0.72 values, while the DBN model shows the second best model with RMSE: 98 KW.hr/m<sup>2</sup>, MAE: 74.48 KW.hr/m<sup>2</sup>, AIC: 17103, NSE: 0.64 and, KGE: 0.74. The analysis results have the potential to provide information to decision makers and planners, especially in the evaluation of the performance of solar panels, the optimization of agricultural processes, and the development of climate change adaptation strategies.</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 189-204"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453799","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Numerical study on PCM geometries to enhance thermal performance of building envelopes PCM几何形状对提高建筑围护结构热工性能的数值研究

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-09-26 DOI: 10.1016/j.jer.2025.09.014

EL Mers EL Mahdi , Aicha Fathi , Ouhammou Badr , Naaim Soufyane , Mihi Miriam , EL Merabet Youssef , Daouchi Brahim , Aggour Mohammed , Zouhair Sadoune

This research focuses on the thermal performance of construction walls with PCMs in a moderate climate, the main aim being the improvement of passive temperature control and a decrease in energy resource requirements. The main aim is to investigate the transient response of temperature distribution and PCM state change (in terms of liquid fraction) within the wall. The transient numerical simulation was performed using ANSYS Fluent based on enthalpy method to capture latent heat storage and subsequent release due to phase-change. Characteristic parameters under investigation are the temperature distribution across the wall layers as well as the liquid fraction of the PCM during charging and discharging. Different PCM categories and melting points were considered to give evidence regarding advantageous thermal performance. A high-resolution simulation setup for the boundary conditions of a moderate climate is employed, and what is new is that both thermal and phase transition physics are considered. Characteristic parameters under investigation are the temperature distribution across the wall layers as well as the liquid fraction of the PCM during charging and discharging. Different PCM categories and melting points were considered to give evidence regarding advantageous thermal performance. A high-resolution simulation setup for the boundary conditions of a moderate climate is employed, and what is new is that both thermal and phase transition physics are considered. One of the main contributions of this work is to explore the PCM geometry optimization requirements. Simulation results demonstrate that optimized PCM geometries significantly accelerate thermal response, improve phase change uniformity, and enhance heat distribution across the wall

本研究的重点是在温和的气候条件下建筑墙体的热工性能，主要目的是改善被动式温度控制和减少能源需求。主要目的是研究壁内温度分布和PCM状态变化（以液体分数表示）的瞬态响应。利用ANSYS Fluent软件基于焓法进行瞬态数值模拟，捕捉相变引起的潜热蓄积和后续释放。所研究的特征参数是在充电和放电过程中跨壁层的温度分布以及PCM的液体分数。不同的PCM类别和熔点被认为是有利的热性能的证据。采用高分辨率模拟装置模拟温和气候的边界条件，新颖之处在于考虑了热物理和相变物理。所研究的特征参数是在充电和放电过程中跨壁层的温度分布以及PCM的液体分数。不同的PCM类别和熔点被认为是有利的热性能的证据。采用高分辨率模拟装置模拟温和气候的边界条件，新颖之处在于考虑了热物理和相变物理。这项工作的主要贡献之一是探索了PCM几何优化要求。仿真结果表明，优化后的PCM几何形状显著加快了热响应速度，改善了相变均匀性，增强了壁面上的热分布

{"title":"Numerical study on PCM geometries to enhance thermal performance of building envelopes","authors":"EL Mers EL Mahdi , Aicha Fathi , Ouhammou Badr , Naaim Soufyane , Mihi Miriam , EL Merabet Youssef , Daouchi Brahim , Aggour Mohammed , Zouhair Sadoune","doi":"10.1016/j.jer.2025.09.014","DOIUrl":"10.1016/j.jer.2025.09.014","url":null,"abstract":"<div><div>This research focuses on the thermal performance of construction walls with PCMs in a moderate climate, the main aim being the improvement of passive temperature control and a decrease in energy resource requirements. The main aim is to investigate the transient response of temperature distribution and PCM state change (in terms of liquid fraction) within the wall. The transient numerical simulation was performed using ANSYS Fluent based on enthalpy method to capture latent heat storage and subsequent release due to phase-change. Characteristic parameters under investigation are the temperature distribution across the wall layers as well as the liquid fraction of the PCM during charging and discharging. Different PCM categories and melting points were considered to give evidence regarding advantageous thermal performance. A high-resolution simulation setup for the boundary conditions of a moderate climate is employed, and what is new is that both thermal and phase transition physics are considered. Characteristic parameters under investigation are the temperature distribution across the wall layers as well as the liquid fraction of the PCM during charging and discharging. Different PCM categories and melting points were considered to give evidence regarding advantageous thermal performance. A high-resolution simulation setup for the boundary conditions of a moderate climate is employed, and what is new is that both thermal and phase transition physics are considered. One of the main contributions of this work is to explore the PCM geometry optimization requirements. Simulation results demonstrate that optimized PCM geometries significantly accelerate thermal response, improve phase change uniformity, and enhance heat distribution across the wall</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 536-553"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Neural-network-based model for predicting asphaltene stability in crude oils 基于神经网络的原油沥青质稳定性预测模型

IF 2.2 4区工程技术 Q3 ENGINEERING, MULTIDISCIPLINARY

Journal of Engineering Research

Pub Date : 2026-03-01 Epub Date: 2025-09-19 DOI: 10.1016/j.jer.2025.08.022

Mohammed S. Alhajeri , Yousef E. Alshamlan , Mohammed M. Alajmi , Ali Elkamel

Asphaltene deposition poses a major challenge in oilfield operations, leading to pipeline blockages, reduced reservoir permeability, and eventual declines in production output. When asphaltenes become destabilized in crude oil, they precipitate and form aggregates, leading to significant deposition. Therefore, monitoring the stability of asphaltenes in crude oil is crucial to prevent the exacerbation of these problems. The weight subfractions are a critical input for calculating any asphaltene stability index; however, obtaining them involves costly and time-consuming experimental procedures. In this research, a statistical machine-learning-based artificial neural network model is proposed to predict this critical parameter at a wide range of operational conditions, with satisfactory accuracy. An extended dataset of over 200 experimental data points was collected from the literature and used to train and test the proposed model. The results demonstrated that the proposed model performed very well versus several commonly used asphaltene prediction models: colloidal instability index, colloidal stability index, asphaltene-to-resins-ratio, and Stankiewicz stability plot. This suggests that the proposed model can be a valuable tool for providing input to various asphaltene modeling tasks.

沥青质沉积是油田作业的主要挑战，会导致管道堵塞、储层渗透率降低，最终导致产量下降。当沥青质在原油中变得不稳定时，它们会沉淀并形成聚集体，导致严重的沉积。因此，监测原油中沥青质的稳定性对于防止这些问题的恶化至关重要。重量分量是计算任何沥青质稳定性指数的关键输入；然而，获得它们涉及昂贵和耗时的实验过程。在本研究中，提出了一种基于统计机器学习的人工神经网络模型，用于在广泛的操作条件下预测这一关键参数，并具有令人满意的精度。从文献中收集了超过200个实验数据点的扩展数据集，并用于训练和测试所提出的模型。结果表明，与几种常用的沥青质预测模型（胶体不稳定性指数、胶体稳定性指数、沥青质与树脂比和Stankiewicz稳定性图）相比，该模型的预测效果非常好。这表明所提出的模型可以成为为各种沥青质建模任务提供输入的有价值的工具。

{"title":"Neural-network-based model for predicting asphaltene stability in crude oils","authors":"Mohammed S. Alhajeri , Yousef E. Alshamlan , Mohammed M. Alajmi , Ali Elkamel","doi":"10.1016/j.jer.2025.08.022","DOIUrl":"10.1016/j.jer.2025.08.022","url":null,"abstract":"<div><div>Asphaltene deposition poses a major challenge in oilfield operations, leading to pipeline blockages, reduced reservoir permeability, and eventual declines in production output. When asphaltenes become destabilized in crude oil, they precipitate and form aggregates, leading to significant deposition. Therefore, monitoring the stability of asphaltenes in crude oil is crucial to prevent the exacerbation of these problems. The weight subfractions are a critical input for calculating any asphaltene stability index; however, obtaining them involves costly and time-consuming experimental procedures. In this research, a statistical machine-learning-based artificial neural network model is proposed to predict this critical parameter at a wide range of operational conditions, with satisfactory accuracy. An extended dataset of over 200 experimental data points was collected from the literature and used to train and test the proposed model. The results demonstrated that the proposed model performed very well versus several commonly used asphaltene prediction models: colloidal instability index, colloidal stability index, asphaltene-to-resins-ratio, and Stankiewicz stability plot. This suggests that the proposed model can be a valuable tool for providing input to various asphaltene modeling tasks.</div></div>","PeriodicalId":48803,"journal":{"name":"Journal of Engineering Research","volume":"14 1","pages":"Pages 20-33"},"PeriodicalIF":2.2,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147453832","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0