Applied Thermal Engineering最新文献_第10页

Computationally efficient nonlinear model predictive control with integral action for ultra-supercritical coal fired units under wide load operation with various disturbances 具有积分作用的超超临界燃煤机组大负荷工况非线性预测控制

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-23 DOI: 10.1016/j.applthermaleng.2026.129947

He Fan , Hengrui Zhang , Yongzhen Wang , Xianyong Peng , Wen Sheng , Li Sun

Ultra-supercritical (USC) coal fired units are required to improve operational flexibility, in order to absorb more renewable energy generation into power grid. However, strong nonlinearity and various disturbances deteriorate the control performance of coordinated control system (CCS) severely. To this end, this work proposes a computationally efficient nonlinear model predictive control (NMPC) method with integral action. Firstly, successive linearization (SL) is used to obtain linear predictive model at each sampling interval, and control action can be calculated online in an explicit form to promote the calculational efficiency. Then integral action is combined with the NMPC to reject various disturbances containing real measure noises, and the detailed procedure for parameter tuning is presented to meet different requirements on tracking performance and variation rate of control action. Lastly, stability analysis and simulation tests are performed to validate its effectiveness. Simulation results reveal that the proposed method has excellent computational efficiency, load tracking and anti-disturbance performances under wide load range from 30% to 100% rated load compared with the NMPC-SL method, constant MPC, conventional proportional-integral-derivative control and neural network generalized predictive control. Its computational efficiency increases by 87% compared with NMPC methods using quadratic programming. Besides, the designed CCS owns the satisfactory root mean square errors, namely, 0.453 MPa, 4.223 kJ/kg and 0.725 MW, and the mean absolute relative error of unit load decreases by at least 80% compared with other control strategies. Therefore, the proposed method can provide reference for improving operational flexibility and anti-disturbance performances of USC units.

超超临界（USC）燃煤机组需要提高运行灵活性，以便将更多的可再生能源发电吸收到电网中。然而，较强的非线性和各种干扰严重影响了协调控制系统的控制性能。为此，本文提出了一种计算效率高的积分作用非线性模型预测控制（NMPC）方法。首先，采用逐次线性化（SL）方法在每个采样区间获得线性预测模型，并在线显式计算控制动作，提高计算效率；然后将积分作用与NMPC相结合，抑制含有实测噪声的各种干扰，并给出了参数整定的详细步骤，以满足对跟踪性能和控制作用变化率的不同要求。最后进行了稳定性分析和仿真试验，验证了该方法的有效性。仿真结果表明，与NMPC-SL方法、恒MPC方法、传统的比例-积分-导数控制和神经网络广义预测控制相比，该方法在30% ~ 100%额定负荷范围内具有优异的计算效率、负载跟踪和抗干扰性能。与使用二次规划的NMPC方法相比，其计算效率提高了87%。设计的CCS具有满意的均方根误差，分别为0.453 MPa、4.223 kJ/kg和0.725 MW，单位负荷的平均绝对相对误差比其他控制策略降低了至少80%。因此，该方法可为提高超超临界机组的操作灵活性和抗干扰性能提供参考。

{"title":"Computationally efficient nonlinear model predictive control with integral action for ultra-supercritical coal fired units under wide load operation with various disturbances","authors":"He Fan , Hengrui Zhang , Yongzhen Wang , Xianyong Peng , Wen Sheng , Li Sun","doi":"10.1016/j.applthermaleng.2026.129947","DOIUrl":"10.1016/j.applthermaleng.2026.129947","url":null,"abstract":"<div><div>Ultra-supercritical (USC) coal fired units are required to improve operational flexibility, in order to absorb more renewable energy generation into power grid. However, strong nonlinearity and various disturbances deteriorate the control performance of coordinated control system (CCS) severely. To this end, this work proposes a computationally efficient nonlinear model predictive control (NMPC) method with integral action. Firstly, successive linearization (SL) is used to obtain linear predictive model at each sampling interval, and control action can be calculated online in an explicit form to promote the calculational efficiency. Then integral action is combined with the NMPC to reject various disturbances containing real measure noises, and the detailed procedure for parameter tuning is presented to meet different requirements on tracking performance and variation rate of control action. Lastly, stability analysis and simulation tests are performed to validate its effectiveness. Simulation results reveal that the proposed method has excellent computational efficiency, load tracking and anti-disturbance performances under wide load range from 30% to 100% rated load compared with the NMPC-SL method, constant MPC, conventional proportional-integral-derivative control and neural network generalized predictive control. Its computational efficiency increases by 87% compared with NMPC methods using quadratic programming. Besides, the designed CCS owns the satisfactory root mean square errors, namely, 0.453 MPa, 4.223 kJ/kg and 0.725 MW, and the mean absolute relative error of unit load decreases by at least 80% compared with other control strategies. Therefore, the proposed method can provide reference for improving operational flexibility and anti-disturbance performances of USC units.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129947"},"PeriodicalIF":6.9,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146074686","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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 optimization of geometries of a domestic biomass pellet stove 国产生物质颗粒炉几何形状的数值优化

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-23 DOI: 10.1016/j.applthermaleng.2026.129938

Yanzhi Chen , Tao Zhang , Jingyong Cai , Zhengrong Shi , Tao Ma , Xinli Lu , Zhichao Wang , Zhou Zheng , Jiayu Xiao , Yalin Chen

Domestic biomass pellet stoves are expected to provide low-carbon heating while simultaneously meeting efficiency and emission limits; however, most prior research targets industrial boilers, leaving stove-specific modeling and multi-parameter couplings insufficiently resolved. This study develops a full-scale CFD–DPM model of a pellet stove and validates it against a physical platform using three-dimensional temperature fields and outlet CO/NO_x measurements. Based on the validated model, a systematic parametric study investigates the impact of furnace geometry (length, height, and width) and fuel type on combustion, heat transfer, and emissions. The results indicate that moderate elongation improves burnout, whereas excessive length introduces dilution and pressure losses. Intermediate height enhances temperature retention and effective residence time, which initially reduces CO concentration before an increase at excessive heights due to reduced temperature and incomplete combustion. However, the increase in furnace length also extends the high-temperature residence time, which can accelerate NOx formation. An intermediate width maximizes centrally located lateral mixing while avoiding near-wall cold regions. Fuel chemistry exerts a dominant influence: coconut shells provide the most favorable trade-off (high completeness with low CO and relatively low NO_x), beech wood pellets offer balanced and compliant behavior, wood chips minimize NO_x but tend to elevate CO, and wood pellets perform the weakest overall. For the geometry considered, combining L = 225 mm, H = 520 mm, and W = 290 mm with coconut shells (or beech wood pellets when coconut shells are unavailable) achieves the optimal trade-off among effective residence time, temperature retention, and pressure drop, while satisfying the Italian five-star standard. The results provide quantitative guidance for the design and operation of domestic biomass pellet stoves.

国内生物质颗粒炉有望提供低碳供暖，同时满足效率和排放限制；然而，大多数先前的研究都是针对工业锅炉的，没有充分解决特定炉子的建模和多参数耦合问题。本研究开发了颗粒炉的全尺寸CFD-DPM模型，并使用三维温度场和出口CO/NOx测量在物理平台上对其进行了验证。基于验证模型，系统参数研究了炉的几何形状（长度、高度和宽度）和燃料类型对燃烧、传热和排放的影响。结果表明，适度伸长率可以改善燃尽，而过大的长度则会导致稀释和压力损失。中等高度增强了温度保持和有效停留时间，这首先降低了CO浓度，然后在过高高度由于温度降低和不完全燃烧而增加。然而，炉长的增加也延长了高温停留时间，这可以加速NOx的形成。中间宽度最大限度地提高了位于中央的横向混合，同时避免了近壁冷区域。燃料化学发挥着主导作用：椰子壳提供了最有利的权衡（高完整性，低CO和相对较低的NOx），山毛榉木屑提供了平衡和合规的行为，木屑将NOx降至最低，但往往会提高CO，木屑的整体表现最差。考虑到几何形状，将L = 225 mm, H = 520 mm和W = 290 mm与椰子壳（或当椰子壳不可用时山毛榉木颗粒）相结合，在有效停留时间，温度保持和压降之间实现了最佳权衡，同时满足意大利五星级标准。研究结果为国产生物质颗粒炉的设计和运行提供了定量指导。

{"title":"Numerical optimization of geometries of a domestic biomass pellet stove","authors":"Yanzhi Chen , Tao Zhang , Jingyong Cai , Zhengrong Shi , Tao Ma , Xinli Lu , Zhichao Wang , Zhou Zheng , Jiayu Xiao , Yalin Chen","doi":"10.1016/j.applthermaleng.2026.129938","DOIUrl":"10.1016/j.applthermaleng.2026.129938","url":null,"abstract":"<div><div>Domestic biomass pellet stoves are expected to provide low-carbon heating while simultaneously meeting efficiency and emission limits; however, most prior research targets industrial boilers, leaving stove-specific modeling and multi-parameter couplings insufficiently resolved. This study develops a full-scale CFD–DPM model of a pellet stove and validates it against a physical platform using three-dimensional temperature fields and outlet CO/NO<sub>x</sub> measurements. Based on the validated model, a systematic parametric study investigates the impact of furnace geometry (length, height, and width) and fuel type on combustion, heat transfer, and emissions. The results indicate that moderate elongation improves burnout, whereas excessive length introduces dilution and pressure losses. Intermediate height enhances temperature retention and effective residence time, which initially reduces CO concentration before an increase at excessive heights due to reduced temperature and incomplete combustion. However, the increase in furnace length also extends the high-temperature residence time, which can accelerate NOx formation. An intermediate width maximizes centrally located lateral mixing while avoiding near-wall cold regions. Fuel chemistry exerts a dominant influence: coconut shells provide the most favorable trade-off (high completeness with low CO and relatively low NO<sub>x</sub>), beech wood pellets offer balanced and compliant behavior, wood chips minimize NO<sub>x</sub> but tend to elevate CO, and wood pellets perform the weakest overall. For the geometry considered, combining L = 225 mm, H = 520 mm, and W = 290 mm with coconut shells (or beech wood pellets when coconut shells are unavailable) achieves the optimal trade-off among effective residence time, temperature retention, and pressure drop, while satisfying the Italian five-star standard. The results provide quantitative guidance for the design and operation of domestic biomass pellet stoves.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129938"},"PeriodicalIF":6.9,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146074436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Dynamic thermal characteristics of the air reservoir in a compressed air energy storage system during air charge and discharge processes 压缩空气储能系统中储气罐在空气充放过程中的动态热特性

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-23 DOI: 10.1016/j.applthermaleng.2026.129891

Hongyi Liu , Lei Liu , Ming Shi , Kaiyuan Jin , Yao Zhao

The rapid increase of renewable energy across the global power sector is driving unprecedented demand for large-scale energy storage. Compressed air energy storage becomes a potential solution due to its large capacity, long lifespan, and low operating cost. The transient heat transfer process within the air reservoir strongly affects the durability of the reservoir and the energy conversion efficiency of the whole system. In this study, a two-dimensional axisymmetric numerical model was developed to investigate the dynamic thermal behavior of a compressed air reservoir. The model was validated using test data from the Huntorf power plant and then used to simulate an artificial reservoir designed recently. A comprehensive parametric study was performed to investigate the effects of air flow rates, inlet temperature, and reservoir height to diameter ratio on the dynamic heat transfer characteristics during charging and discharging processes. The results reveal that a high-temperature zone could form at the bottom of the reservoir when the air is being injected, resulting in a large temperature gradient along the reservoir wall. Higher inlet temperatures can increase average temperature, pressure and stored enthalpy within the reservoir but have little effect on changing the peak temperature of the high temperature zone. Variation of air flow rates for both charging and discharging processes slightly influences the temperature profile but could not significantly reduce the large temperature difference within the reservoir, either. Nevertheless, the study finds that the temperature uniformity of the air reservoir is primarily dominated by the reservoir height-to-diameter ratio. When this ratio is set to be around 5, the overall temperature field becomes almost uniform, and the peak temperature within the reservoir is significantly reduced. The developed model and the reported findings provide a critical basis for future studies, which could further analyze the thermal-mechanical performance of the compressed air energy storage reservoirs and develop the optimal design with least thermal stress issues and highest mechanical stability. These efforts will accelerate the worldwide development of compressed air energy storage technology and further increase the share of renewables in global power grids.

全球电力部门可再生能源的快速增长推动了对大规模储能的前所未有的需求。压缩空气储能因其容量大、寿命长、运行成本低而成为一种潜在的解决方案。储气罐内部的瞬态换热过程对储气罐的耐久性和整个系统的能量转换效率有很大影响。本文建立了一个二维轴对称数值模型来研究压缩空气储层的动态热行为。利用Huntorf电厂的试验数据对模型进行了验证，并应用于近期设计的人工水库的模拟。通过综合参数研究，探讨了进气流量、进气温度和储层高径比对充放过程动态换热特性的影响。结果表明：空气注入会在储层底部形成高温区，沿储层壁面产生较大的温度梯度；较高的入口温度可以提高库内平均温度、压力和储存焓，但对高温区峰值温度的变化影响不大。充放风流量的变化对库内温度分布的影响较小，但也不能显著减小库内较大的温差。但研究发现，储层温度均匀性主要受储层高径比的影响。当该比值设置为5左右时，整体温度场趋于均匀，库内峰值温度明显降低。所建立的模型和研究结果为进一步分析压缩空气蓄能器的热力学性能，制定热应力问题最少、力学稳定性最高的优化设计提供了重要的基础。这些努力将加速压缩空气储能技术的全球发展，并进一步提高可再生能源在全球电网中的份额。

{"title":"Dynamic thermal characteristics of the air reservoir in a compressed air energy storage system during air charge and discharge processes","authors":"Hongyi Liu , Lei Liu , Ming Shi , Kaiyuan Jin , Yao Zhao","doi":"10.1016/j.applthermaleng.2026.129891","DOIUrl":"10.1016/j.applthermaleng.2026.129891","url":null,"abstract":"<div><div>The rapid increase of renewable energy across the global power sector is driving unprecedented demand for large-scale energy storage. Compressed air energy storage becomes a potential solution due to its large capacity, long lifespan, and low operating cost. The transient heat transfer process within the air reservoir strongly affects the durability of the reservoir and the energy conversion efficiency of the whole system. In this study, a two-dimensional axisymmetric numerical model was developed to investigate the dynamic thermal behavior of a compressed air reservoir. The model was validated using test data from the Huntorf power plant and then used to simulate an artificial reservoir designed recently. A comprehensive parametric study was performed to investigate the effects of air flow rates, inlet temperature, and reservoir height to diameter ratio on the dynamic heat transfer characteristics during charging and discharging processes. The results reveal that a high-temperature zone could form at the bottom of the reservoir when the air is being injected, resulting in a large temperature gradient along the reservoir wall. Higher inlet temperatures can increase average temperature, pressure and stored enthalpy within the reservoir but have little effect on changing the peak temperature of the high temperature zone. Variation of air flow rates for both charging and discharging processes slightly influences the temperature profile but could not significantly reduce the large temperature difference within the reservoir, either. Nevertheless, the study finds that the temperature uniformity of the air reservoir is primarily dominated by the reservoir height-to-diameter ratio. When this ratio is set to be around 5, the overall temperature field becomes almost uniform, and the peak temperature within the reservoir is significantly reduced. The developed model and the reported findings provide a critical basis for future studies, which could further analyze the thermal-mechanical performance of the compressed air energy storage reservoirs and develop the optimal design with least thermal stress issues and highest mechanical stability. These efforts will accelerate the worldwide development of compressed air energy storage technology and further increase the share of renewables in global power grids.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129891"},"PeriodicalIF":6.9,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146074758","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Research on attention-based fault diagnosis and multi-parameter joint optimization of CO2 heat pump system CO2热泵系统基于注意力的故障诊断及多参数联合优化研究

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-23 DOI: 10.1016/j.applthermaleng.2026.129942

Yabin Guo , Congcong Du , Xin Liu , Xingtao Zhang , Zunlong Jin

Heating, ventilation and air conditioning (HVAC) systems in buildings are susceptible to various faults, compromising operational efficiency and leading to unnecessary energy consumption. Therefore, it holds significant engineering importance to carry out fault diagnosis research on the system. As a deep learning method, attention mechanism has gained growing attention and application recently. This study proposed an attention-based fault diagnosis model for carbon dioxide heat pump (CHP) systems, which was trained and validated with a real-world fault database obtained from a transcritical CHP system. The database encompassed four common fault types, including water flow reduction fault, evaporator fouling fault, electronic expansion valve stuck fault and fan fault. Each fault type had multiple severity levels, with fault data collected under stable operating conditions. The impact of key hyperparameters on the diagnostic performance of the model was systematically evaluated, including batch size, learning rate, weight decay, patience, and label smoothing factor. To address the complex coupling between these parameters, the Bayesian optimization algorithm was employed for simultaneous multi-parameter tuning. The optimized model demonstrated significant improvements in both diagnostic accuracy and operational robustness compared to the unoptimized model and current mainstream fault diagnosis models. Specifically, the accuracy, precision and recall of the optimized model all reached 94.46%, and the F1-score was 94.44%.

建筑物的供暖、通风和空调（HVAC）系统容易出现各种故障，影响运行效率并导致不必要的能源消耗。因此，对系统进行故障诊断研究具有重要的工程意义。注意机制作为一种深度学习方法，近年来得到了越来越多的关注和应用。本文提出了一种基于注意力的二氧化碳热泵系统故障诊断模型，并利用一个跨临界热电联产系统的真实故障数据库对该模型进行了训练和验证。该数据库包含四种常见的故障类型，包括水流减少故障、蒸发器结垢故障、电子膨胀阀卡死故障和风扇故障。每种故障类型都有多个严重级别，在稳定运行条件下收集故障数据。系统地评估了关键超参数对模型诊断性能的影响，包括批大小、学习率、权重衰减、耐心和标签平滑因子。针对这些参数之间的复杂耦合，采用贝叶斯优化算法进行多参数同步整定。与未优化模型和当前主流故障诊断模型相比，优化后的模型在诊断精度和运行鲁棒性方面均有显著提高。其中，优化模型的准确率、精密度和召回率均达到94.46%，f1得分为94.44%。

{"title":"Research on attention-based fault diagnosis and multi-parameter joint optimization of CO2 heat pump system","authors":"Yabin Guo , Congcong Du , Xin Liu , Xingtao Zhang , Zunlong Jin","doi":"10.1016/j.applthermaleng.2026.129942","DOIUrl":"10.1016/j.applthermaleng.2026.129942","url":null,"abstract":"<div><div>Heating, ventilation and air conditioning (HVAC) systems in buildings are susceptible to various faults, compromising operational efficiency and leading to unnecessary energy consumption. Therefore, it holds significant engineering importance to carry out fault diagnosis research on the system. As a deep learning method, attention mechanism has gained growing attention and application recently. This study proposed an attention-based fault diagnosis model for carbon dioxide heat pump (CHP) systems, which was trained and validated with a real-world fault database obtained from a transcritical CHP system. The database encompassed four common fault types, including water flow reduction fault, evaporator fouling fault, electronic expansion valve stuck fault and fan fault. Each fault type had multiple severity levels, with fault data collected under stable operating conditions. The impact of key hyperparameters on the diagnostic performance of the model was systematically evaluated, including batch size, learning rate, weight decay, patience, and label smoothing factor. To address the complex coupling between these parameters, the Bayesian optimization algorithm was employed for simultaneous multi-parameter tuning. The optimized model demonstrated significant improvements in both diagnostic accuracy and operational robustness compared to the unoptimized model and current mainstream fault diagnosis models. Specifically, the accuracy, precision and recall of the optimized model all reached 94.46%, and the F1-score was 94.44%.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129942"},"PeriodicalIF":6.9,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146074500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Investigations on an RBFNN-based optimization method for the two-phase flow microchannel thermal management system of multi-chip power electronic devices 基于rbfnn的多芯片电力电子器件两相流微通道热管理系统优化方法研究

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-23 DOI: 10.1016/j.applthermaleng.2026.129936

Chenzhen Ji , Zhuo Zhang , Zhengpeng Chen , Yong Chen , Fei Duan

Efficient thermal management in high-power electronic devices requires cooling strategies that can simultaneously suppress temperature rise without introducing excessive flow resistance. Traditional computational fluid dynamics (CFD)-based microchannel optimization design requires a large amount of computational resources and time, especially for the cases involving two-phase flow heat transfer. This study presents a comprehensive multi-objective optimization framework that combines CFD with advanced surrogate model approximation to accelerate the design process of two-phase flow microchannels, improving thermal performance and reducing energy consumption in multi-chip power electronic devices. This approach integrates Latin Hypercube Sampling (LHS), Radial Basis Function Neural Network (RBFNN), and Non-dominated Sorting Genetic Algorithm II (NSGA-II) for sample selection, model training, and objective optimization. The multiple optimization objectives are to minimize the average chip temperature (

T_{ave}

) while simultaneously minimizing the pressure drop (

Δ P

) in the cooling channel by changing the design variables of microchannel width, fin thickness, and cold plate thickness. The RBFNN, validated through k-fold cross-validation, achieves superior predictive accuracy of heat transfer performance compared with traditional correlation-based surrogate models, with mean relative errors of only 0.07% for

T_{ave}

and 3.1% for

Δ P

. Optimization through the NSGA-II algorithm generates Pareto fronts that capture the fundamental trade-off between cooling enhancement and flow resistance, resulting in three solutions: minimum

T_{ave}

, minimum

Δ P

, and a balanced solution that reduces energy consumption while improving the reliability of electronic devices. The Figure of Merit is calculated and compared for a comprehensive performance analysis of microchannel thermal management designs. Furthermore, a sensitivity analysis is conducted based on the existing dataset to definitively identify the order of importance of various factors affecting temperature, pressure drop, as well as the Nusselt number and friction factor. The proposed framework provides a highly efficient AI-based multi-objective optimization method for two-phase flow microchannel thermal management systems compared with fully resolved three-dimensional CFD simulations, thereby advancing the development of next-generation energy-efficient power electronics cooling technologies.

在高功率电子设备中，高效的热管理需要能够同时抑制温度上升而不引入过多流动阻力的冷却策略。传统的基于计算流体力学（CFD）的微通道优化设计需要大量的计算资源和时间，特别是对于涉及两相流传热的情况。本研究提出了一种综合的多目标优化框架，将CFD与先进的替代模型近似相结合，以加速两相流微通道的设计过程，提高多芯片电力电子器件的热性能，降低能耗。该方法集成了拉丁超立方采样（LHS）、径向基函数神经网络（RBFNN）和非支配排序遗传算法II (NSGA-II)，用于样本选择、模型训练和目标优化。通过改变微通道宽度、翅片厚度和冷板厚度等设计变量，实现最小化芯片平均温度（Tave）的同时最小化冷却通道内的压降（ΔP）的多重优化目标。通过k-fold交叉验证验证的RBFNN与传统的基于相关性的替代模型相比，具有更高的传热性能预测精度，Tave的平均相对误差仅为0.07%，ΔP的平均相对误差为3.1%。通过NSGA-II算法的优化生成了帕累托前沿，捕获了冷却增强和流动阻力之间的基本权衡，从而产生了三种解决方案：最小Tave，最小ΔP，以及在提高电子设备可靠性的同时降低能耗的平衡解决方案。计算并比较了微通道热管理设计的综合性能分析。此外，基于现有数据集进行敏感性分析，明确确定影响温度、压降以及努塞尔数和摩擦因数的各种因素的重要程度。与全分辨率三维CFD模拟相比，该框架为两相流微通道热管理系统提供了一种高效的基于人工智能的多目标优化方法，从而推动了下一代节能电力电子冷却技术的发展。

{"title":"Investigations on an RBFNN-based optimization method for the two-phase flow microchannel thermal management system of multi-chip power electronic devices","authors":"Chenzhen Ji , Zhuo Zhang , Zhengpeng Chen , Yong Chen , Fei Duan","doi":"10.1016/j.applthermaleng.2026.129936","DOIUrl":"10.1016/j.applthermaleng.2026.129936","url":null,"abstract":"<div><div>Efficient thermal management in high-power electronic devices requires cooling strategies that can simultaneously suppress temperature rise without introducing excessive flow resistance. Traditional computational fluid dynamics (CFD)-based microchannel optimization design requires a large amount of computational resources and time, especially for the cases involving two-phase flow heat transfer. This study presents a comprehensive multi-objective optimization framework that combines CFD with advanced surrogate model approximation to accelerate the design process of two-phase flow microchannels, improving thermal performance and reducing energy consumption in multi-chip power electronic devices. This approach integrates Latin Hypercube Sampling (LHS), Radial Basis Function Neural Network (RBFNN), and Non-dominated Sorting Genetic Algorithm II (NSGA-II) for sample selection, model training, and objective optimization. The multiple optimization objectives are to minimize the average chip temperature (<span><math><msub><mrow><mi>T</mi></mrow><mrow><mi>ave</mi></mrow></msub></math></span>) while simultaneously minimizing the pressure drop (<span><math><mrow><mi>Δ</mi><mi>P</mi></mrow></math></span>) in the cooling channel by changing the design variables of microchannel width, fin thickness, and cold plate thickness. The RBFNN, validated through k-fold cross-validation, achieves superior predictive accuracy of heat transfer performance compared with traditional correlation-based surrogate models, with mean relative errors of only 0.07% for <span><math><msub><mrow><mi>T</mi></mrow><mrow><mi>ave</mi></mrow></msub></math></span> and 3.1% for <span><math><mrow><mi>Δ</mi><mi>P</mi></mrow></math></span>. Optimization through the NSGA-II algorithm generates Pareto fronts that capture the fundamental trade-off between cooling enhancement and flow resistance, resulting in three solutions: minimum <span><math><msub><mrow><mi>T</mi></mrow><mrow><mi>ave</mi></mrow></msub></math></span>, minimum <span><math><mrow><mi>Δ</mi><mi>P</mi></mrow></math></span>, and a balanced solution that reduces energy consumption while improving the reliability of electronic devices. The Figure of Merit is calculated and compared for a comprehensive performance analysis of microchannel thermal management designs. Furthermore, a sensitivity analysis is conducted based on the existing dataset to definitively identify the order of importance of various factors affecting temperature, pressure drop, as well as the Nusselt number and friction factor. The proposed framework provides a highly efficient AI-based multi-objective optimization method for two-phase flow microchannel thermal management systems compared with fully resolved three-dimensional CFD simulations, thereby advancing the development of next-generation energy-efficient power electronics cooling technologies.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129936"},"PeriodicalIF":6.9,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146045182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Transparent photovoltaic solar chimney systems for desert sustainability: modelling, energy benefits, and potential for rainfall enhancement 沙漠可持续发展的透明光伏太阳能烟囱系统：建模、能源效益和增强降雨的潜力

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-22 DOI: 10.1016/j.applthermaleng.2026.129875

Jing Nie , Tong-Zheng Guo , Li-Yao-Min Nie , Jin-Chen Xu , Jing Jia

This study develops and evaluates a large-scale transparent photovoltaic solar chimney power plant (PV-SCPP) as a sustainable solution for desert ecological restoration and climate regulation. A comprehensive mathematical model and numerical simulation platform were created to examine the system's dual benefits: renewable energy generation and precipitation enhancement. A new Rainfall Factor is introduced to quantify the additional precipitation triggered by the system through an evaporation-condensation process. The results show that the transparent PV-SCPP produces about thirteen times the energy output of a conventional solar chimney, while also fostering favorable microclimatic conditions for rainfall. To enhance precipitation, the system requires chimney heights of at least 600 m, solar irradiation above 600 W/m², relative humidity above 25%, and top wind speeds below 10.5 m/s. A comparative analysis of the Ulan Buh and Badain Jaran Deserts shows that the Ulan Buh Desert is better suited for condensation formation and rainfall promotion. Beyond its energy performance, this study emphasizes the potential of transparent PV-SCPP technology to aid in desert climate regulation, ecological restoration, and long-term regional sustainability.

本研究开发并评估了一种大型透明光伏太阳能烟囱电厂（PV-SCPP），作为沙漠生态恢复和气候调节的可持续解决方案。建立了一个综合的数学模型和数值模拟平台来检验该系统的双重效益：可再生能源发电和降水增强。引入了一个新的降雨因子来量化系统通过蒸发-冷凝过程引发的额外降水。结果表明，透明的PV-SCPP产生的能量输出大约是传统太阳能烟囱的13倍，同时也为降雨提供了有利的小气候条件。为了增强降水，系统要求烟囱高度不小于600m，太阳辐照度大于600w /m2，相对湿度大于25%，最大风速小于10.5 m/s。乌兰布和沙漠与巴丹吉林沙漠的对比分析表明，乌兰布和沙漠更适合凝结形成和降雨促进。除了能源性能之外，本研究还强调了透明PV-SCPP技术在沙漠气候调节、生态恢复和长期区域可持续性方面的潜力。

{"title":"Transparent photovoltaic solar chimney systems for desert sustainability: modelling, energy benefits, and potential for rainfall enhancement","authors":"Jing Nie , Tong-Zheng Guo , Li-Yao-Min Nie , Jin-Chen Xu , Jing Jia","doi":"10.1016/j.applthermaleng.2026.129875","DOIUrl":"10.1016/j.applthermaleng.2026.129875","url":null,"abstract":"<div><div>This study develops and evaluates a large-scale transparent photovoltaic solar chimney power plant (PV-SCPP) as a sustainable solution for desert ecological restoration and climate regulation. A comprehensive mathematical model and numerical simulation platform were created to examine the system's dual benefits: renewable energy generation and precipitation enhancement. A new Rainfall Factor is introduced to quantify the additional precipitation triggered by the system through an evaporation-condensation process. The results show that the transparent PV-SCPP produces about thirteen times the energy output of a conventional solar chimney, while also fostering favorable microclimatic conditions for rainfall. To enhance precipitation, the system requires chimney heights of at least 600 m, solar irradiation above 600 W/m<sup>2</sup>, relative humidity above 25%, and top wind speeds below 10.5 m/s. A comparative analysis of the Ulan Buh and Badain Jaran Deserts shows that the Ulan Buh Desert is better suited for condensation formation and rainfall promotion. Beyond its energy performance, this study emphasizes the potential of transparent PV-SCPP technology to aid in desert climate regulation, ecological restoration, and long-term regional sustainability.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129875"},"PeriodicalIF":6.9,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146074501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Copper inverse opal composite wicking structures for high-performance thermal management in compact electronic devices 用于紧凑型电子设备中高性能热管理的铜反蛋白石复合芯结构

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-22 DOI: 10.1016/j.applthermaleng.2026.129894

Shiwei Zhang , Ruicheng Wang , Yiming Li , Tong Sun , Fangqiong Luo , Zhijie Li , Jingjing Bai , Wanghuai Xu , Yong Tang

With the continuous development of electronic devices toward higher performance, higher integration, and miniaturization, ultra-thin vapor chambers (UTVCs) have been increasingly widely used in efficient thermal management of confined spaces. However, in extremely thin spaces, the wicking structure has become a key factor restricting the further performance breakthrough of UTVCs. To address this challenge, this study proposes a novel wicking structure based on copper inverse opal (CIO), which significantly enhances capillary performance without increasing thickness. The mesh with copper inverse opal (MCIO) prepared by this process exhibits a capillary coefficient of 18.96 mm·s^−0.5, which is 85.9% and 534.1% higher than that of the traditional wicking structure and the inverse opal (IO) structure, respectively. Furthermore, the thickness of the UTVC integrated with the MCIO wicking structure is only 0.29 mm. Experimental investigations reveal that the optimal filling ratio of this UTVC is 40%. Under this filling ratio and an inclination angle of 90°, the UTVC achieves a maximum heat load of 6 W, a minimum total thermal resistance of only 0.85 °C/W, and a maximum effective thermal conductivity of 13,707.57 W/(m·K). The maximum effective thermal conductivity of the MCIO-UTVC sample reaches approximately 206.5% of that recorded for the baseline M-UTVC. In addition, the infrared experimental results show that compared with the solid copper plate, the UTVC exhibits excellent heat dissipation performance and temperature uniformity. The research results indicate that the copper inverse opal composite process has great potential in enhancing the heat transfer performance of ultra-thin phase-change heat transfer devices.

随着电子器件向高性能、高集成度和小型化的不断发展，超薄蒸汽室（utvc）在密闭空间的高效热管理中得到了越来越广泛的应用。然而，在极薄的空间中，排芯结构已成为制约utvc进一步突破性能的关键因素。为了解决这一挑战，本研究提出了一种基于铜反蛋白石（CIO）的新型芯结构，该结构在不增加厚度的情况下显著提高了毛细管性能。该工艺制备的铜反蛋白石（MCIO）网状物的毛细系数为18.96 mm·s−0.5，分别比传统吸芯结构和反蛋白石（IO）结构提高了85.9%和534.1%。与MCIO排芯结构相结合的UTVC厚度仅为0.29 mm。实验研究表明，该UTVC的最佳填充率为40%。在此填充比和90°倾角下，UTVC的最大热负荷为6 W，最小总热阻仅为0.85°C/W，最大有效导热系数为13707.57 W/（m·K）。MCIO-UTVC样品的最大有效热导率约为基线M-UTVC记录的206.5%。此外，红外实验结果表明，与固体铜板相比，UTVC具有良好的散热性能和温度均匀性。研究结果表明，铜反蛋白石复合工艺在提高超薄相变传热器件的传热性能方面具有很大的潜力。

{"title":"Copper inverse opal composite wicking structures for high-performance thermal management in compact electronic devices","authors":"Shiwei Zhang , Ruicheng Wang , Yiming Li , Tong Sun , Fangqiong Luo , Zhijie Li , Jingjing Bai , Wanghuai Xu , Yong Tang","doi":"10.1016/j.applthermaleng.2026.129894","DOIUrl":"10.1016/j.applthermaleng.2026.129894","url":null,"abstract":"<div><div>With the continuous development of electronic devices toward higher performance, higher integration, and miniaturization, ultra-thin vapor chambers (UTVCs) have been increasingly widely used in efficient thermal management of confined spaces. However, in extremely thin spaces, the wicking structure has become a key factor restricting the further performance breakthrough of UTVCs. To address this challenge, this study proposes a novel wicking structure based on copper inverse opal (CIO), which significantly enhances capillary performance without increasing thickness. The mesh with copper inverse opal (MCIO) prepared by this process exhibits a capillary coefficient of 18.96 mm·s<sup>−0.5</sup>, which is 85.9% and 534.1% higher than that of the traditional wicking structure and the inverse opal (IO) structure, respectively. Furthermore, the thickness of the UTVC integrated with the MCIO wicking structure is only 0.29 mm. Experimental investigations reveal that the optimal filling ratio of this UTVC is 40%. Under this filling ratio and an inclination angle of 90°, the UTVC achieves a maximum heat load of 6 W, a minimum total thermal resistance of only 0.85 °C/W, and a maximum effective thermal conductivity of 13,707.57 W/(m·K). The maximum effective thermal conductivity of the MCIO-UTVC sample reaches approximately 206.5% of that recorded for the baseline M-UTVC. In addition, the infrared experimental results show that compared with the solid copper plate, the UTVC exhibits excellent heat dissipation performance and temperature uniformity. The research results indicate that the copper inverse opal composite process has great potential in enhancing the heat transfer performance of ultra-thin phase-change heat transfer devices.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129894"},"PeriodicalIF":6.9,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146045178","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Lightweight thermal management of permanent magnet synchronous machines for extreme power density operation 轻量级的热管理永磁同步电机的极端功率密度操作

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-22 DOI: 10.1016/j.applthermaleng.2026.129890

Boyu Chen, Jiwei Cao, Liyi Li, Haoyu Chen, Sen Yan

Taking the drive motor of a hydrogen-oxygen electric pump as a representative example of short-duration ultra-high power density machines, this study investigates lightweight thermal management techniques under extreme power density operating conditions. An evaluation method for the limiting output power is first proposed to determine the theoretical maximum output power of the motor under given constraints. However, the copper loss associated with this limiting operating point exceeds the allowable level of the machine. To address this issue, a novel cooling configuration is introduced, in which cryogenic channels are embedded in the stator slots and supplied with a low-temperature coolant. The proposed structure operates in two modes: immersion cooling and flow cooling. In the former, the vaporization latent heat of a quiescent coolant is utilized to remove heat, while in the latter, forced coolant flow enhances convective heat transfer. To clarify the underlying mechanisms, an energized solenoid is initially adopted as a simplified test object to investigate and compare the heat transfer performance of the two cooling modes, and immersion cooling is further validated experimentally. Subsequently, an 80 kW prototype motor is used as a test platform to conduct both cryogenic flow and immersion cooling experiments, thereby demonstrating the feasibility and effectiveness of the proposed lightweight thermal management approach.

本研究以短寿命超高功率密度机器的氢氧电泵驱动电机为代表，探讨了在极端功率密度工况下的轻量化热管理技术。首先提出了一种极限输出功率的评估方法，以确定给定约束条件下电机的理论最大输出功率。然而，与这个极限工作点相关的铜损耗超过了机器的允许水平。为了解决这个问题，引入了一种新的冷却结构，在定子槽中嵌入低温通道，并提供低温冷却剂。所提出的结构工作在两种模式：浸没冷却和流动冷却。在前者中，静止冷却剂的汽化潜热被用来去除热量，而在后者中，强制冷却剂流动增强对流换热。为了弄清其背后的机理，首先采用通电螺线管作为简化测试对象，研究和比较两种冷却方式的换热性能，并进一步通过实验验证浸入式冷却。随后，以一台80 kW的原型电机作为测试平台，进行了低温流动和浸入式冷却实验，从而验证了所提出的轻量化热管理方法的可行性和有效性。

{"title":"Lightweight thermal management of permanent magnet synchronous machines for extreme power density operation","authors":"Boyu Chen, Jiwei Cao, Liyi Li, Haoyu Chen, Sen Yan","doi":"10.1016/j.applthermaleng.2026.129890","DOIUrl":"10.1016/j.applthermaleng.2026.129890","url":null,"abstract":"<div><div>Taking the drive motor of a hydrogen-oxygen electric pump as a representative example of short-duration ultra-high power density machines, this study investigates lightweight thermal management techniques under extreme power density operating conditions. An evaluation method for the limiting output power is first proposed to determine the theoretical maximum output power of the motor under given constraints. However, the copper loss associated with this limiting operating point exceeds the allowable level of the machine. To address this issue, a novel cooling configuration is introduced, in which cryogenic channels are embedded in the stator slots and supplied with a low-temperature coolant. The proposed structure operates in two modes: immersion cooling and flow cooling. In the former, the vaporization latent heat of a quiescent coolant is utilized to remove heat, while in the latter, forced coolant flow enhances convective heat transfer. To clarify the underlying mechanisms, an energized solenoid is initially adopted as a simplified test object to investigate and compare the heat transfer performance of the two cooling modes, and immersion cooling is further validated experimentally. Subsequently, an 80 kW prototype motor is used as a test platform to conduct both cryogenic flow and immersion cooling experiments, thereby demonstrating the feasibility and effectiveness of the proposed lightweight thermal management approach.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129890"},"PeriodicalIF":6.9,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146074493","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Forecasting-driven multi-objective optimal configuration of a wind-solar-molten salt hybrid energy system for the alumina digestion process 氧化铝溶出过程中风-太阳能-熔盐混合能源系统的多目标优化配置

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-22 DOI: 10.1016/j.applthermaleng.2026.129870

Liang Tang , Kaiyue Li , Hongwei Wang , Jiying Liu , Zuming Liu

In response to the “ Carbon Peaking and Carbon Neutrality goals “ strategic objectives, the energy supply system for the alumina digestion process is shifting toward a green energy-dominated model. This study develops a hybrid green energy-molten salt synergistic system for supplying energy to the alumina digestion process, based on wind and solar power forecasting. The autoregressive integrated moving average-long short-term memory (ARIMA-LSTM) hybrid prediction model provides highly accurate meteorological forecasts, achieving an average coefficient of determination (R²) value greater than 97.9%. The non-dominated sorting genetic algorithm III (NSGA-III) is a multi-objective optimization algorithm addressing the complex trade-offs between three objective functions by combining the Pareto dominance principle with fuzzy membership functions. In addition, four typical configuration schemes are designed for a 168-h weekly multi-objective optimization study. The complementarity index (CI), coefficient of variation (CV), and information entropy (H) are used as evaluation metrics. The obtained results show that the full-equipment collaborative configuration scheme has remarkable cross-seasonal stable energy supply characteristics, achieving an optimal balance between economy, environmental performance, and energy efficiency. The minimum weekly operating cost, minimum weekly carbon emissions, and maximum green energy consumption rate are 7.73 million CHY, 71.23 t, and 99.73%, respectively. The configuration with molten salt thermal storage achieves the optimal CV and H, effectively stabilizing the wind-solar fluctuations, reducing the grid power purchase volatility by 62–81%, and increasing the stability of energy supply by 71–86%. Furthermore, the synergistic complementarity between wind turbines and photovoltaic systems effectively mitigates the temporal and intensity limitations of single energy sources.

为响应“碳峰值和碳中和目标”的战略目标，氧化铝消化过程的能源供应系统正在向以绿色能源为主导的模式转变。本研究开发了一种基于风能和太阳能预测的混合绿色能源-熔盐协同系统，为氧化铝消化过程提供能源。自回归综合移动平均-长短期记忆（ARIMA-LSTM）混合预测模型预报精度高，平均决定系数（R2）大于97.9%。非支配排序遗传算法III （non- dominant sorting genetic algorithm III, NSGA-III）是将Pareto支配原则与模糊隶属函数相结合，解决三个目标函数之间复杂权衡的多目标优化算法。此外，设计了4种典型配置方案，进行了每周168 h的多目标优化研究。以互补性指数（CI）、变异系数（CV）和信息熵(H)作为评价指标。结果表明，全设备协同配置方案具有显著的跨季节稳定供能特性，实现了经济性、环保性和能效之间的最佳平衡。周最小运行成本为773万CHY，周最小碳排放为71.23 t，周最大绿色能源消耗率为99.73%。采用熔盐蓄热配置实现了最优CV和H值，有效稳定了风能-太阳能波动，使电网购电波动率降低62-81%，能源供应稳定性提高71-86%。此外，风力涡轮机和光伏系统之间的协同互补性有效地减轻了单一能源的时间和强度限制。

{"title":"Forecasting-driven multi-objective optimal configuration of a wind-solar-molten salt hybrid energy system for the alumina digestion process","authors":"Liang Tang , Kaiyue Li , Hongwei Wang , Jiying Liu , Zuming Liu","doi":"10.1016/j.applthermaleng.2026.129870","DOIUrl":"10.1016/j.applthermaleng.2026.129870","url":null,"abstract":"<div><div>In response to the “ Carbon Peaking and Carbon Neutrality goals “ strategic objectives, the energy supply system for the alumina digestion process is shifting toward a green energy-dominated model. This study develops a hybrid green energy-molten salt synergistic system for supplying energy to the alumina digestion process, based on wind and solar power forecasting. The autoregressive integrated moving average-long short-term memory (ARIMA-LSTM) hybrid prediction model provides highly accurate meteorological forecasts, achieving an average coefficient of determination (R<sup>2</sup>) value greater than 97.9%. The non-dominated sorting genetic algorithm III (NSGA-III) is a multi-objective optimization algorithm addressing the complex trade-offs between three objective functions by combining the Pareto dominance principle with fuzzy membership functions. In addition, four typical configuration schemes are designed for a 168-h weekly multi-objective optimization study. The complementarity index (CI), coefficient of variation (CV), and information entropy (H) are used as evaluation metrics. The obtained results show that the full-equipment collaborative configuration scheme has remarkable cross-seasonal stable energy supply characteristics, achieving an optimal balance between economy, environmental performance, and energy efficiency. The minimum weekly operating cost, minimum weekly carbon emissions, and maximum green energy consumption rate are 7.73 million CHY, 71.23 t, and 99.73%, respectively. The configuration with molten salt thermal storage achieves the optimal CV and H, effectively stabilizing the wind-solar fluctuations, reducing the grid power purchase volatility by 62–81%, and increasing the stability of energy supply by 71–86%. Furthermore, the synergistic complementarity between wind turbines and photovoltaic systems effectively mitigates the temporal and intensity limitations of single energy sources.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"289 ","pages":"Article 129870"},"PeriodicalIF":6.9,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024093","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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

Optimisation study of solar-coupled combined heat and power for crude oil export terminals 原油出口终端太阳能耦合热电联产优化研究

IF 6.9 2区工程技术 Q2 ENERGY & FUELS

Applied Thermal Engineering

Pub Date : 2026-01-22 DOI: 10.1016/j.applthermaleng.2026.129787

Liang Tian, Fuxing Zhao, Jiachao Wu

Addressing the issues of high energy consumption, high carbon emissions, and significant load fluctuations at oilfield crude export terminals, this paper investigates a solar cogeneration system. Its core lies in employing coupled molten salt thermal storage technology to resolve the contradictions between energy intermittency and seasonality. Using Aspen Plus software, a collection-storage-Rankine cycle model was constructed for different working fluids. Embedding thermodynamic constraints via a physical information neural network, the study developed an ANN-NSGA-II multi-objective collaborative optimization framework. This achieved Pareto optimisation of the system's thermodynamic energy efficiency, economic benefits, and environmental performance. The multi-objective optimisation comparison results indicate that the optimal

{RD}_{2}

system (a combined heat and power configuration utilizing molten salt for direct collection and storage, with a steam Rankine cycle for power generation) based on a steam Rankine cycle achieves a maximum thermal efficiency of 55.43%, with an internal rate of return (IRR) of 11.23% and annual CO₂ emissions reductions of 931.36 tonnes. Sensitivity analysis indicates that a power/heat coupling diverter ratio exceeding 0.4 ensures positive net electricity generation and sustained carbon reduction growth. Increasing the diversion ratio diminishes IRR, permitting dynamic operational optimisation based on real-time meteorological conditions and grid peak-shaving demands. The molten salt thermal storage system enables effective intraday energy management. It facilitates intelligent operational scheduling based on electricity price signals: generating power at full capacity during periods of high solar irradiance and high electricity prices, while storing heat during low-price periods. This strategy, optimized via the NSGA-II algorithm within a multi-dimensional framework considering solar radiation, electricity pricing, and load patterns, significantly reduces daily reliance on external heat supplementation and grid electricity purchases, demonstrating the engineering feasibility of the proposed system for managing daily energy imbalances.

This work establishes a replicable, optimized technical paradigm for solar-assisted decarbonization of industrial energy systems, offering a viable pathway for the large-scale substitution of fossil fuels in the oil and gas sector.

针对油田原油出口终端高能耗、高碳排放、负荷波动大的问题，对太阳能热电联产系统进行了研究。其核心在于采用耦合熔盐蓄热技术，解决能源间歇性与季节性的矛盾。利用Aspen Plus软件，建立了不同工质的收集-储存-朗肯循环模型。通过物理信息神经网络嵌入热力学约束，开发了ANN-NSGA-II多目标协同优化框架。这实现了系统热力学能源效率、经济效益和环境性能的帕累托优化。多目标优化对比结果表明，以蒸汽朗肯循环为基础的RD2系统（采用熔盐直接收集和储存，采用蒸汽朗肯循环发电的热电联产结构）的最优热效率为55.43%，内部收益率（IRR）为11.23%，年二氧化碳排放量减少931.36吨。敏感性分析表明，功率/热耦合分流比超过0.4，可确保净发电量为正，碳减排持续增长。增加导流比可以降低IRR，允许基于实时气象条件和电网调峰需求的动态操作优化。熔盐蓄热系统可实现有效的日间能源管理。实现基于电价信号的智能调度，在太阳辐照度高、电价高时满负荷发电，在电价低时蓄热。该策略通过NSGA-II算法在考虑太阳辐射、电价和负荷模式的多维框架内进行优化，显著降低了日常对外部供热和电网购电的依赖，证明了所提出的系统在管理日常能源失衡方面的工程可行性。这项工作为太阳能辅助工业能源系统脱碳建立了一个可复制的、优化的技术范例，为石油和天然气部门大规模替代化石燃料提供了一条可行的途径。

{"title":"Optimisation study of solar-coupled combined heat and power for crude oil export terminals","authors":"Liang Tian, Fuxing Zhao, Jiachao Wu","doi":"10.1016/j.applthermaleng.2026.129787","DOIUrl":"10.1016/j.applthermaleng.2026.129787","url":null,"abstract":"<div><div>Addressing the issues of high energy consumption, high carbon emissions, and significant load fluctuations at oilfield crude export terminals, this paper investigates a solar cogeneration system. Its core lies in employing coupled molten salt thermal storage technology to resolve the contradictions between energy intermittency and seasonality. Using Aspen Plus software, a collection-storage-Rankine cycle model was constructed for different working fluids. Embedding thermodynamic constraints via a physical information neural network, the study developed an ANN-NSGA-II multi-objective collaborative optimization framework. This achieved Pareto optimisation of the system's thermodynamic energy efficiency, economic benefits, and environmental performance. The multi-objective optimisation comparison results indicate that the optimal <span><math><msub><mi>RD</mi><mn>2</mn></msub></math></span> system (a combined heat and power configuration utilizing molten salt for direct collection and storage, with a steam Rankine cycle for power generation) based on a steam Rankine cycle achieves a maximum thermal efficiency of 55.43%, with an internal rate of return (IRR) of 11.23% and annual CO₂ emissions reductions of 931.36 tonnes. Sensitivity analysis indicates that a power/heat coupling diverter ratio exceeding 0.4 ensures positive net electricity generation and sustained carbon reduction growth. Increasing the diversion ratio diminishes IRR, permitting dynamic operational optimisation based on real-time meteorological conditions and grid peak-shaving demands. The molten salt thermal storage system enables effective intraday energy management. It facilitates intelligent operational scheduling based on electricity price signals: generating power at full capacity during periods of high solar irradiance and high electricity prices, while storing heat during low-price periods. This strategy, optimized via the NSGA-II algorithm within a multi-dimensional framework considering solar radiation, electricity pricing, and load patterns, significantly reduces daily reliance on external heat supplementation and grid electricity purchases, demonstrating the engineering feasibility of the proposed system for managing daily energy imbalances.</div><div>This work establishes a replicable, optimized technical paradigm for solar-assisted decarbonization of industrial energy systems, offering a viable pathway for the large-scale substitution of fossil fuels in the oil and gas sector.</div></div>","PeriodicalId":8201,"journal":{"name":"Applied Thermal Engineering","volume":"290 ","pages":"Article 129787"},"PeriodicalIF":6.9,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146122542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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