Building and Environment最新文献_第7页

Photosynthetically active radiation of photovoltaic-green wall: An experimental and simulation study in hot summer and cold winter area 夏热冬冷地区光伏绿墙光合有效辐射的实验与模拟研究

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-21 DOI: 10.1016/j.buildenv.2025.114160

Mengmeng Dai, Yang He, Yongfa Diao, Jianlin Liu

In the context of mitigating the urban heat island effect and achieving carbon-neutral goals, integrating photovoltaics with greenery on building façades has emerged as a promising solution for sustainable urban development. However, façade spaces remain underexplored, especially the effects of photovoltaic shading on the photosynthetically active radiation on vegetation surfaces. In this study, field measurements were conducted to analyze the spatiotemporal distribution of solar radiation and photosynthetically active radiation in a photovoltaic-green wall system in a hot summer and cold winter area. For the south-facing wall, maximum incident radiation on the plant surface between the front surface of the photovoltaic module and the photovoltaic array is 371.9 W/m² and 365.6 W/m², respectively, while below the photovoltaic modules, it drops to only 225.6 W/m², exhibiting significant spatial heterogeneity. A validated anisotropic solar radiation model that accounts for photovoltaic shading achieved an R² > 0.84 against measured data. Using the least-squares method, a linear relationship between solar radiation and photosynthetically active radiation was established. Based on this model, a sensitivity analysis quantified the effect of key design parameters on daily light integral. Moreover, yearly simulation results reveal spatiotemporal daily light integral differences across different orientations: the upper region of the south wall meets the high-light requirements of plants for more than 21% of the year, while the north wall remains at low-light levels for long periods. The findings are expected to provide a reference for optimizing the allocation of solar-thermal resources in the integrated design of building facades with photovoltaics and greenery.

在缓解城市热岛效应和实现碳中和目标的背景下，在建筑立面上整合光伏与绿色植物已经成为可持续城市发展的一个有希望的解决方案。然而，遮阳对植被表面光合有效辐射的影响仍未得到充分研究。在夏热冬冷地区，通过野外实测分析了光伏-绿墙系统中太阳辐射和光合有效辐射的时空分布。对南向壁面，光伏组件正面与光伏阵列之间的植物表面最大入射辐射分别为371.9 W/m²和365.6 W/m²，而光伏组件下方的植物表面最大入射辐射仅为225.6 W/m²，空间异质性显著。考虑光伏遮阳的经验证的各向异性太阳辐射模型与实测数据相比获得了R²>； 0.84。利用最小二乘法，建立了太阳辐射与光合有效辐射之间的线性关系。基于该模型，进行了灵敏度分析，量化了关键设计参数对日光照积分的影响。此外，年度模拟结果揭示了不同方向上的时空日光照积分差异：南墙上部区域满足植物全年21%以上的高光照需求，而北墙长期处于低光照水平。研究结果有望为建筑立面与光伏、绿化一体化设计中的光热资源优化配置提供参考。

{"title":"Photosynthetically active radiation of photovoltaic-green wall: An experimental and simulation study in hot summer and cold winter area","authors":"Mengmeng Dai, Yang He, Yongfa Diao, Jianlin Liu","doi":"10.1016/j.buildenv.2025.114160","DOIUrl":"10.1016/j.buildenv.2025.114160","url":null,"abstract":"<div><div>In the context of mitigating the urban heat island effect and achieving carbon-neutral goals, integrating photovoltaics with greenery on building façades has emerged as a promising solution for sustainable urban development. However, façade spaces remain underexplored, especially the effects of photovoltaic shading on the photosynthetically active radiation on vegetation surfaces. In this study, field measurements were conducted to analyze the spatiotemporal distribution of solar radiation and photosynthetically active radiation in a photovoltaic-green wall system in a hot summer and cold winter area. For the south-facing wall, maximum incident radiation on the plant surface between the front surface of the photovoltaic module and the photovoltaic array is 371.9 W/m² and 365.6 W/m², respectively, while below the photovoltaic modules, it drops to only 225.6 W/m², exhibiting significant spatial heterogeneity. A validated anisotropic solar radiation model that accounts for photovoltaic shading achieved an R² > 0.84 against measured data. Using the least-squares method, a linear relationship between solar radiation and photosynthetically active radiation was established. Based on this model, a sensitivity analysis quantified the effect of key design parameters on daily light integral. Moreover, yearly simulation results reveal spatiotemporal daily light integral differences across different orientations: the upper region of the south wall meets the high-light requirements of plants for more than 21% of the year, while the north wall remains at low-light levels for long periods. The findings are expected to provide a reference for optimizing the allocation of solar-thermal resources in the integrated design of building facades with photovoltaics and greenery.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114160"},"PeriodicalIF":7.6,"publicationDate":"2025-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

The impact of different air distribution and extraction strategies on the mitigation of airborne infection risk in classroom settings 不同的空气分布和提取策略对减轻教室环境中空气传播感染风险的影响

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-20 DOI: 10.1016/j.buildenv.2025.114158

Muhammad Farhan Ejaz , Simo Kilpeläinen , Panu Mustakallio , Risto Kosonen

Mitigating airborne infection risk in classrooms requires effective and reliable ventilation solutions. In this study, a full-scale classroom with eighteen pupils was modeled, and three possible locations of an infected person were considered. Mixing, wall-attached downward, and displacement supply systems were evaluated. Single- and multi-point ceiling exhausts (with one, two, or four outlets) were tested for mixing and wall-attached systems. A canopy-based local exhaust above the occupied zone was analyzed for wall-attached and displacement systems. Moreover, a mixing system scenario with 25 % higher airflow (ACH) and a single exhaust was included for comparison. The validated CFD (computational fluid dynamics) model was utilized to analyze airflow patterns, contaminant concentrations, and infection probability. The results show that the performance of high airflow rate and multi-point extraction strategies improved conditions for only specific infector locations. Since the infector’s location is unknown in practice, these solutions may not be considered robust. However, the occupant-focused canopy exhaust system performed consistently well for all simulated infector locations. In the best-case scenario, the canopy system with displacement ventilation reduced the average contaminant concentration in the occupied zone by approximately 45 % compared to the mixing system with a single exhaust. It also reduced the infection risk more effectively than other strategies, allowing at most two occupants to be in the high-risk category. The findings suggest that targeted local extraction is a more reliable and effective approach for minimizing the risk of airborne infection, offering a consistent solution regardless of the infector’s location.

减轻教室空气传播感染风险需要有效可靠的通风解决方案。在这项研究中，模拟了一个有18名学生的全尺寸教室，并考虑了感染者的三个可能位置。对混合、贴壁向下和置换供应系统进行了评价。测试了单点和多点天花板排气口（有一个、两个或四个出口）用于混合和壁挂式系统。分析了壁挂式和置换式系统在占位区上方基于顶篷的局部排气。此外，为了进行比较，还包括了具有25%高气流（ACH）和单一排气的混合系统场景。验证的CFD（计算流体动力学）模型用于分析气流模式、污染物浓度和感染概率。结果表明，大风量和多点提取策略的性能改善了特定感染部位的条件。由于感染者的位置在实践中是未知的，这些解决方案可能不被认为是可靠的。然而，以乘员为中心的顶篷排气系统在所有模拟感染位置都表现良好。在最好的情况下，与单一排气的混合系统相比，带置换通风的顶篷系统将占用区域的平均污染物浓度降低了约45%。它还比其他策略更有效地降低了感染风险，最多允许两名居住者属于高风险类别。研究结果表明，有针对性的局部拔牙是一种更可靠、更有效的方法，可以最大限度地降低空气传播感染的风险，无论感染者在哪里，都能提供一致的解决方案。

{"title":"The impact of different air distribution and extraction strategies on the mitigation of airborne infection risk in classroom settings","authors":"Muhammad Farhan Ejaz , Simo Kilpeläinen , Panu Mustakallio , Risto Kosonen","doi":"10.1016/j.buildenv.2025.114158","DOIUrl":"10.1016/j.buildenv.2025.114158","url":null,"abstract":"<div><div>Mitigating airborne infection risk in classrooms requires effective and reliable ventilation solutions. In this study, a full-scale classroom with eighteen pupils was modeled, and three possible locations of an infected person were considered. Mixing, wall-attached downward, and displacement supply systems were evaluated. Single- and multi-point ceiling exhausts (with one, two, or four outlets) were tested for mixing and wall-attached systems. A canopy-based local exhaust above the occupied zone was analyzed for wall-attached and displacement systems. Moreover, a mixing system scenario with 25 % higher airflow (ACH) and a single exhaust was included for comparison. The validated CFD (computational fluid dynamics) model was utilized to analyze airflow patterns, contaminant concentrations, and infection probability. The results show that the performance of high airflow rate and multi-point extraction strategies improved conditions for only specific infector locations. Since the infector’s location is unknown in practice, these solutions may not be considered robust. However, the occupant-focused canopy exhaust system performed consistently well for all simulated infector locations. In the best-case scenario, the canopy system with displacement ventilation reduced the average contaminant concentration in the occupied zone by approximately 45 % compared to the mixing system with a single exhaust. It also reduced the infection risk more effectively than other strategies, allowing at most two occupants to be in the high-risk category. The findings suggest that targeted local extraction is a more reliable and effective approach for minimizing the risk of airborne infection, offering a consistent solution regardless of the infector’s location.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114158"},"PeriodicalIF":7.6,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838408","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Uncovering biases in standard occupancy assumptions using urban-scale mobile signaling data 使用城市规模移动信号数据揭示标准占用假设中的偏差

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-20 DOI: 10.1016/j.buildenv.2025.114161

Shuyi Chen , Yihu Zhang , Jun Xiao , Ziliang Wei , Yangyang Wang , Yixing Chen , Hao Zhou , Yang Geng , Borong Lin

Accurate representation of building occupancy is essential for energy demand estimation and urban energy system planning. However, most studies still rely on standard reference occupancy patterns from design codes, which neither reflect recent societal shifts nor capture the complexity required for urban building energy modeling (UBEM). To address these challenges, this study conducts a comprehensive investigation of building occupancy using urban-scale mobile signaling data (MSD). By integrating MSD with building attributes and behavioral knowledge, we develop a downscaling framework that generates hourly building occupancy profiles (persons/m²) and quantifies the biases in standard reference patterns. Applied to more than 200,000 buildings in Beijing, the framework reveals the actual occupancy dynamics across nine building types. The MSD-based profiles achieve 50–86 % higher accuracy than standard references, with an RMSE of 0.003–0.005 when validated against on-site counts. Results reveal substantial discrepancies in widely used standard reference occupancy patterns particularly during weekday daytime: residential occupancy is underestimated by 46–93 %, whereas non-residential occupancy is overestimated by 158–277 %. These deviations propagate into energy simulations, leading to deviations ranging from -15.1 % to +21.8 % in annual end-use consumption and -71.7 % to +36.6 % in hourly electricity use under hot weather. The occupancy biases likely stem from evolving building use patterns and the limits of building-level standards in capturing urban-scale uncertainty. This study underscores the need for updated, scenario-specific reference values to improve occupancy assumptions in both building- and urban-scale energy modeling.

建筑物占用率的准确表示对于能源需求估计和城市能源系统规划至关重要。然而，大多数研究仍然依赖于设计规范中的标准参考占用模式，这些模式既不能反映最近的社会变化，也不能反映城市建筑能源建模（UBEM）所需的复杂性。为了应对这些挑战，本研究利用城市规模的移动信号数据（MSD）对建筑物占用情况进行了全面调查。通过将MSD与建筑属性和行为知识相结合，我们开发了一个降尺度框架，该框架可以生成每小时的建筑占用概况（人/平方米），并量化标准参考模式中的偏差。该框架应用于北京20多万幢建筑，揭示了九种建筑类型的实际占用动态。基于msd的配置文件比标准参考文件的准确度高50 - 86%，当对现场计数进行验证时，RMSE为0.003-0.005。结果显示，广泛使用的标准参考入住率模式存在巨大差异，特别是在工作日白天：住宅入住率被低估了46 - 93%，而非住宅入住率被高估了158 - 277%。这些偏差传播到能源模拟中，导致年最终用电量的偏差范围为- 15.1%至+ 21.8%，炎热天气下每小时用电量的偏差范围为- 71.7%至+ 36.6%。占用偏差可能源于不断发展的建筑使用模式和建筑层面的标准在捕捉城市规模不确定性方面的局限性。本研究强调需要更新特定场景的参考值，以改进建筑和城市尺度能源模型中的占用假设。

{"title":"Uncovering biases in standard occupancy assumptions using urban-scale mobile signaling data","authors":"Shuyi Chen , Yihu Zhang , Jun Xiao , Ziliang Wei , Yangyang Wang , Yixing Chen , Hao Zhou , Yang Geng , Borong Lin","doi":"10.1016/j.buildenv.2025.114161","DOIUrl":"10.1016/j.buildenv.2025.114161","url":null,"abstract":"<div><div>Accurate representation of building occupancy is essential for energy demand estimation and urban energy system planning. However, most studies still rely on standard reference occupancy patterns from design codes, which neither reflect recent societal shifts nor capture the complexity required for urban building energy modeling (UBEM). To address these challenges, this study conducts a comprehensive investigation of building occupancy using urban-scale mobile signaling data (MSD). By integrating MSD with building attributes and behavioral knowledge, we develop a downscaling framework that generates hourly building occupancy profiles (persons/m<sup>2</sup>) and quantifies the biases in standard reference patterns. Applied to more than 200,000 buildings in Beijing, the framework reveals the actual occupancy dynamics across nine building types. The MSD-based profiles achieve 50–86 % higher accuracy than standard references, with an RMSE of 0.003–0.005 when validated against on-site counts. Results reveal substantial discrepancies in widely used standard reference occupancy patterns particularly during weekday daytime: residential occupancy is underestimated by 46–93 %, whereas non-residential occupancy is overestimated by 158–277 %. These deviations propagate into energy simulations, leading to deviations ranging from -15.1 % to +21.8 % in annual end-use consumption and -71.7 % to +36.6 % in hourly electricity use under hot weather. The occupancy biases likely stem from evolving building use patterns and the limits of building-level standards in capturing urban-scale uncertainty. This study underscores the need for updated, scenario-specific reference values to improve occupancy assumptions in both building- and urban-scale energy modeling.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"291 ","pages":"Article 114161"},"PeriodicalIF":7.6,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145947954","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Collaborative optimization of environmental quality and energy consumption in shared space micro-environments based on machine learning models 基于机器学习模型的共享空间微环境环境质量与能耗协同优化

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-20 DOI: 10.1016/j.buildenv.2025.114153

Yong Cheng , Jiangang Lei , Xue Tian

Current methods for creating differentiated micro-environments in shared spaces effectively balance personalized thermal preferences and energy efficiency but often neglect Indoor Air Quality (IAQ) integration when optimizing supply parameters. This study proposes a framework combining machine learning-based IAQ distribution models and regression-based thermal comfort distribution models to jointly optimize micro-environments. Two key air supply parameters-velocity and temperature-were evaluated across 19 cases, which included six experimental conditions and 13 Computational Fluid Dynamics (CFD) simulations, all conducted under cooling conditions in a room with eight manikins, using CO₂ as a tracer gas. For each case, four different infected source locations were assessed. Jet theory was applied to design cases and calculate the characteristic length (y_b), defined as the vertical distance between the supply jet center and the breathing point. Initial attempts to predict contaminant concentration, intake fraction, and exposure index at target locations using linear and exponential regression proved insufficiently accurate. Instead, four machine learning models-Backpropagation Neural Network, Support Vector Machine, Random Forest, and XGBoost-were trained and hyperparameter-optimized, significantly improving validation accuracy. XGBoost achieved the best performance (validation R² up to 0.97). Using this model, IAQ distributions were predicted for 200 ventilation scenarios derived from prior thermal comfort research. Multi-objective optimization revealed critical trade-offs: lower y_b values improved IAQ except when the source was within the supply jet, while higher y_b values reduced energy consumption. When objectives are equally weighted, the optimal y_b values are approximately 0.25 m for sources within the supply air jet and 0.13 m for other sources.

目前在共享空间中创建差异化微环境的方法有效地平衡了个性化的热偏好和能源效率，但在优化供应参数时往往忽略了室内空气质量（IAQ）的集成。本研究提出了基于机器学习的室内空气质量分布模型和基于回归的热舒适分布模型相结合的框架，共同优化微环境。两个关键的送风参数-速度和温度-在19个案例中进行了评估，其中包括6个实验条件和13个计算流体动力学（CFD）模拟，所有这些都是在有8个人体模型的房间的冷却条件下进行的，使用二氧化碳作为示踪气体。对于每个病例，评估了四个不同的感染源位置。应用射流理论设计实例，计算特征长度（yb），定义为供给射流中心与呼吸点之间的垂直距离。最初尝试使用线性和指数回归预测目标地点的污染物浓度、摄入分数和暴露指数，结果证明不够准确。取而代之的是，四个机器学习模型——反向传播神经网络、支持向量机、随机森林和xgboost——经过训练和超参数优化，显著提高了验证的准确性。XGBoost获得了最佳性能（验证R²高达0.97）。利用该模型预测了200种通风方案的室内空气质量分布。多目标优化揭示了关键的权衡：较低的yb值改善了室内空气质量，除非源在供应喷气机内，而较高的yb值降低了能耗。当物镜的权重相等时，供气射流内的源的最佳yb值约为0.25 m，其他源的最佳yb值约为0.13 m。

{"title":"Collaborative optimization of environmental quality and energy consumption in shared space micro-environments based on machine learning models","authors":"Yong Cheng , Jiangang Lei , Xue Tian","doi":"10.1016/j.buildenv.2025.114153","DOIUrl":"10.1016/j.buildenv.2025.114153","url":null,"abstract":"<div><div>Current methods for creating differentiated micro-environments in shared spaces effectively balance personalized thermal preferences and energy efficiency but often neglect Indoor Air Quality (IAQ) integration when optimizing supply parameters. This study proposes a framework combining machine learning-based IAQ distribution models and regression-based thermal comfort distribution models to jointly optimize micro-environments. Two key air supply parameters-velocity and temperature-were evaluated across 19 cases, which included six experimental conditions and 13 Computational Fluid Dynamics (CFD) simulations, all conducted under cooling conditions in a room with eight manikins, using CO₂ as a tracer gas. For each case, four different infected source locations were assessed. Jet theory was applied to design cases and calculate the characteristic length (<em>y<sub>b</sub></em>), defined as the vertical distance between the supply jet center and the breathing point. Initial attempts to predict contaminant concentration, intake fraction, and exposure index at target locations using linear and exponential regression proved insufficiently accurate. Instead, four machine learning models-Backpropagation Neural Network, Support Vector Machine, Random Forest, and XGBoost-were trained and hyperparameter-optimized, significantly improving validation accuracy. XGBoost achieved the best performance (validation R² up to 0.97). Using this model, IAQ distributions were predicted for 200 ventilation scenarios derived from prior thermal comfort research. Multi-objective optimization revealed critical trade-offs: lower <em>y<sub>b</sub></em> values improved IAQ except when the source was within the supply jet, while higher <em>y<sub>b</sub></em> values reduced energy consumption. When objectives are equally weighted, the optimal <em>y<sub>b</sub></em> values are approximately 0.25 m for sources within the supply air jet and 0.13 m for other sources.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114153"},"PeriodicalIF":7.6,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838344","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Can ChatGPT-5 assess human perceptions of dynamic urban streetscape transitions? Insights from Hong Kong ChatGPT-5能否评估人类对动态城市街景转变的感知？香港的启示

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-19 DOI: 10.1016/j.buildenv.2025.114155

Mengyuan Yin, Michael C.P. Sing, Esther H.K. Yung

Urban streetscape transitions influence emotions and perceived restorativeness, yet most studies rely on static imagery. This study investigates whether GPT-5 can approximate human perceptions of dynamic transitions, using Hong Kong as a case. We applied a three-stage methodology consisting of alignment, attribution, and interpretation to systematically compare GPT-5 with human evaluations. Four five-minute walking videos were evaluated by 200 participants on affective states using the Pleasure, Arousal, and Dominance (PAD) model, perceived restorativeness (PRS), and subjective impressions. GPT-5 received stitched video frames and performed three tasks: overall scoring, segment-based scoring with feature attribution, and short language-based explanations. Comparisons showed a moderate positive correlation between GPT-5 and human ratings (Pearson r = 0.44, Spearman ρ = 0.43, both p < 0.001). GPT-5 aligned more closely with humans on structural and physical cues (e.g., greenery and density indicators) but showed markedly lower agreement on affective and restorative dimensions. In contrast, the language-based explanations showed both overlap and divergence. Both GPT-5 and humans highlighted greenery, but their interpretations differed in other aspects: humans emphasized openness and experiential cues, whereas GPT-5 focused on signage, buildings, and traffic. The findings highlight how generative AI responds to dynamic environmental transitions, filling a gap left by static-image studies and providing a useful complement to human judgement in high-density cities.

城市街景的过渡影响情绪和感知恢复性，但大多数研究依赖于静态图像。本研究以香港为例，探讨GPT-5是否可以近似人类对动态转换的感知。我们采用了三阶段的方法，包括校准、归因和解释，系统地比较GPT-5与人类评估。200名参与者使用愉悦、觉醒和支配（PAD）模型、感知恢复（PRS）和主观印象来评估四个五分钟的步行视频的情感状态。GPT-5接收拼接的视频帧，并执行三个任务：总体评分，基于特征归属的基于片段的评分，以及基于简短语言的解释。比较显示GPT-5与人类评分之间存在中度正相关（Pearson r = 0.44, Spearman ρ = 0.43, p < 0.001）。GPT-5在结构和物理线索（如绿化和密度指标）上与人类更接近，但在情感和恢复维度上的一致性明显较低。相比之下，基于语言的解释既有重叠，也有差异。GPT-5和人类都强调绿色，但他们的解释在其他方面有所不同：人类强调开放性和经验线索，而GPT-5则侧重于标志、建筑和交通。研究结果强调了生成式人工智能如何响应动态环境变化，填补了静态图像研究留下的空白，并为高密度城市中的人类判断提供了有用的补充。

{"title":"Can ChatGPT-5 assess human perceptions of dynamic urban streetscape transitions? Insights from Hong Kong","authors":"Mengyuan Yin, Michael C.P. Sing, Esther H.K. Yung","doi":"10.1016/j.buildenv.2025.114155","DOIUrl":"10.1016/j.buildenv.2025.114155","url":null,"abstract":"<div><div>Urban streetscape transitions influence emotions and perceived restorativeness, yet most studies rely on static imagery. This study investigates whether GPT-5 can approximate human perceptions of dynamic transitions, using Hong Kong as a case. We applied a three-stage methodology consisting of alignment, attribution, and interpretation to systematically compare GPT-5 with human evaluations. Four five-minute walking videos were evaluated by 200 participants on affective states using the Pleasure, Arousal, and Dominance (PAD) model, perceived restorativeness (PRS), and subjective impressions. GPT-5 received stitched video frames and performed three tasks: overall scoring, segment-based scoring with feature attribution, and short language-based explanations. Comparisons showed a moderate positive correlation between GPT-5 and human ratings (Pearson <em>r</em> = 0.44, Spearman ρ = 0.43, both <em>p</em> < 0.001). GPT-5 aligned more closely with humans on structural and physical cues (e.g., greenery and density indicators) but showed markedly lower agreement on affective and restorative dimensions. In contrast, the language-based explanations showed both overlap and divergence. Both GPT-5 and humans highlighted greenery, but their interpretations differed in other aspects: humans emphasized openness and experiential cues, whereas GPT-5 focused on signage, buildings, and traffic. The findings highlight how generative AI responds to dynamic environmental transitions, filling a gap left by static-image studies and providing a useful complement to human judgement in high-density cities.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114155"},"PeriodicalIF":7.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Paradigm shifts in indoor air quality: Insights from the 2025 Stanford forum on sustainable and healthy buildings 室内空气质量的范式转变：来自2025年斯坦福可持续和健康建筑论坛的见解

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-19 DOI: 10.1016/j.buildenv.2025.114152

Milana Trounce , Darin Anderson , Liam Bates , Seema Bhangar , Rob Bolin , Wenhao Chen , Sabrina Chwalek , Scott Frank , Jason Hartke , Kazukiyo Kumagai , Georgia Lagoudas , Erik Malmstrom , Sean McCrady , Corey Metzger , Aleksander Mikszewski , Daniel Nall , Brendan Owens , John Salas , Steve Taylor , Walt Vernon , Lidia Morawska

This paper presents insights from the 2025 Stanford Forum on Sustainable and Healthy Buildings, co-hosted by Stanford University and Queensland University of Technology. The forum convened experts across government, academia, public health, and industry to examine the feasibility of implementing indoor air quality (IAQ) performance standards aligned with energy efficiency and climate goals. Informed by the recent Science consensus article by Morawska et al. (2024), discussions focused on advancing enforceable IAQ metrics, economic and policy barriers, retrofitting challenges, and certification frameworks. A phased approach—voluntary monitoring, required reporting, and eventual compliance—was widely endorsed. Participants emphasized the need for cost-benefit framing, market incentives, and data transparency. Key outcomes included working groups dedicated to standards and policy approaches, building the economic case and market alignment, pilot projects, and education and knowledge dissemination. This paper summarizes forum findings and offers practical strategies to elevate clean indoor air as essential public health infrastructure, advancing both human and planetary health through healthier, more resilient buildings.

本文介绍了由斯坦福大学和昆士兰科技大学共同主办的2025年斯坦福可持续和健康建筑论坛的见解。该论坛召集了来自政府、学术界、公共卫生和工业界的专家，研究实施符合能源效率和气候目标的室内空气质量（IAQ）绩效标准的可行性。根据Morawska等人（2024）最近发表的《科学共识》（Science consensus）文章，讨论的重点是推进可执行的室内空气质量指标、经济和政策障碍、改造挑战和认证框架。一种分阶段的方法——自愿监督、要求报告和最终遵守——得到了广泛的认可。与会者强调需要成本效益框架、市场激励和数据透明度。主要成果包括专门负责标准和政策方法、建立经济案例和市场协调、试点项目以及教育和知识传播的工作组。本文总结了论坛的研究结果，并提出了切实可行的战略，将清洁的室内空气提升为必不可少的公共卫生基础设施，通过更健康、更有弹性的建筑促进人类和地球的健康。

{"title":"Paradigm shifts in indoor air quality: Insights from the 2025 Stanford forum on sustainable and healthy buildings","authors":"Milana Trounce , Darin Anderson , Liam Bates , Seema Bhangar , Rob Bolin , Wenhao Chen , Sabrina Chwalek , Scott Frank , Jason Hartke , Kazukiyo Kumagai , Georgia Lagoudas , Erik Malmstrom , Sean McCrady , Corey Metzger , Aleksander Mikszewski , Daniel Nall , Brendan Owens , John Salas , Steve Taylor , Walt Vernon , Lidia Morawska","doi":"10.1016/j.buildenv.2025.114152","DOIUrl":"10.1016/j.buildenv.2025.114152","url":null,"abstract":"<div><div>This paper presents insights from the 2025 Stanford Forum on Sustainable and Healthy Buildings, co-hosted by Stanford University and Queensland University of Technology. The forum convened experts across government, academia, public health, and industry to examine the feasibility of implementing indoor air quality (IAQ) performance standards aligned with energy efficiency and climate goals. Informed by the recent Science consensus article by Morawska et al. (2024), discussions focused on advancing enforceable IAQ metrics, economic and policy barriers, retrofitting challenges, and certification frameworks. A phased approach—voluntary monitoring, required reporting, and eventual compliance—was widely endorsed. Participants emphasized the need for cost-benefit framing, market incentives, and data transparency. Key outcomes included working groups dedicated to standards and policy approaches, building the economic case and market alignment, pilot projects, and education and knowledge dissemination. This paper summarizes forum findings and offers practical strategies to elevate clean indoor air as essential public health infrastructure, advancing both human and planetary health through healthier, more resilient buildings.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114152"},"PeriodicalIF":7.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838414","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Using facial expression to predict indoor thermal-acoustic comfort 利用面部表情预测室内热声舒适度

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-19 DOI: 10.1016/j.buildenv.2025.114154

Xin Wen , Qi Meng , Da Yang , Mengmeng Li , Xuejun Hu , Jian Kang

Non-intrusive facial expression measurement has been proven effective in predicting human perceptions in single-domain environments. However, the mechanisms by which this technology influences and predicts comfort under the interactive effects of thermal-acoustic environments have not yet been thoroughly explored. Investigating this area is not only crucial for enhancing prediction accuracy but also for uncovering the patterns of facial expression changes in complex thermal-acoustic environments and their underlying mechanisms affecting comfort. To address this gap, this study conducted experiments in an environmental chamber, involving 6 common indoor temperature levels (18, 20, 22, 24, 26, and 28 °C) and 9 sound pressure levels (35, 40, 45, 50, 55, 60, 65, 70, and 75 dBA). The results showed that facial action units (AUs) can effectively reflect changes in comfort under thermal-acoustic environments, with AU04 and AU14 being core predictive indicators across models. Using a random forest classifier, the proposed model performs well in predicting seven levels of comfort based on facial and thermal–acoustic features. Additionally, the relationship between facial expressions and comfort varied across different thermal-acoustic environments: in comfortable thermal-acoustic environments, more active facial expressions correlated with higher comfort, whereas the opposite was true in harsh environments. This study deepens understanding of how environmental interactions influence facial expressions and comfort, offering insights for multi-domain prediction models, while acknowledging that the use of a controlled laboratory setting and limited demographic scope may constrain generalizability.

非侵入式面部表情测量已被证明在单域环境下预测人类感知的有效性。然而，在热声环境的相互作用下，这项技术影响和预测舒适性的机制尚未得到彻底的探索。研究这一领域不仅对提高预测精度至关重要，而且对揭示复杂热声环境下面部表情变化的模式及其影响舒适度的潜在机制也至关重要。为了解决这一差距，本研究在一个环境室内进行了实验，涉及6个常见的室内温度水平（18、20、22、24、26和28°C）和9个声压水平（35、40、45、50、55、60、65、70和75 dBA）。结果表明，面部动作单元（AUs）能有效反映热声环境下舒适性的变化，其中AU04和AU14是各模型的核心预测指标。使用随机森林分类器，该模型可以很好地预测基于面部和热声特征的七个舒适度水平。此外，面部表情与舒适度之间的关系在不同的热声环境中也有所不同：在舒适的热声环境中，更活跃的面部表情与更高的舒适度相关，而在恶劣的环境中则相反。这项研究加深了对环境相互作用如何影响面部表情和舒适度的理解，为多领域预测模型提供了见解，同时承认使用受控的实验室环境和有限的人口范围可能会限制通用性。

{"title":"Using facial expression to predict indoor thermal-acoustic comfort","authors":"Xin Wen , Qi Meng , Da Yang , Mengmeng Li , Xuejun Hu , Jian Kang","doi":"10.1016/j.buildenv.2025.114154","DOIUrl":"10.1016/j.buildenv.2025.114154","url":null,"abstract":"<div><div>Non-intrusive facial expression measurement has been proven effective in predicting human perceptions in single-domain environments. However, the mechanisms by which this technology influences and predicts comfort under the interactive effects of thermal-acoustic environments have not yet been thoroughly explored. Investigating this area is not only crucial for enhancing prediction accuracy but also for uncovering the patterns of facial expression changes in complex thermal-acoustic environments and their underlying mechanisms affecting comfort. To address this gap, this study conducted experiments in an environmental chamber, involving 6 common indoor temperature levels (18, 20, 22, 24, 26, and 28 °C) and 9 sound pressure levels (35, 40, 45, 50, 55, 60, 65, 70, and 75 dBA). The results showed that facial action units (AUs) can effectively reflect changes in comfort under thermal-acoustic environments, with AU04 and AU14 being core predictive indicators across models. Using a random forest classifier, the proposed model performs well in predicting seven levels of comfort based on facial and thermal–acoustic features. Additionally, the relationship between facial expressions and comfort varied across different thermal-acoustic environments: in comfortable thermal-acoustic environments, more active facial expressions correlated with higher comfort, whereas the opposite was true in harsh environments. This study deepens understanding of how environmental interactions influence facial expressions and comfort, offering insights for multi-domain prediction models, while acknowledging that the use of a controlled laboratory setting and limited demographic scope may constrain generalizability.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114154"},"PeriodicalIF":7.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Structure-property relationships and recyclability of natural fiber insulation 天然纤维绝热材料的结构-性能关系及可回收性

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-19 DOI: 10.1016/j.buildenv.2025.114156

Meghan E. Lamm , Bayleigh M. Loving , Katie Copenhaver , Amber M. Hubbard , Caitlyn M. Clarkson

Thermal insulation can significantly decrease the energy required to maintain internal temperatures within buildings, saving money and decreasing environmental impacts. Many industrially available insulation options utilize petroleum-derived materials, such as polyurethane, fiberglass, and polystyrene, which can be detrimental to the environment by emitting greenhouse gases during production. A sustainable alternative to these synthetic insulations is the use of renewable feedstocks to produce comparable insulation products. In this work, natural fibers such as flax and banana were bound with an epoxy system to produce natural-fiber insulation. Key experimental parameters for this study included evaluating the fiber lengths and processing methods while varying the viscosity of the epoxy binder solvent dissolution. Insulation performance was determined by comparing density, thermal conductivity, resilience, and compressive strength; these parameters were tested to further understand and investigate the structure- property relationships within the system and to understand how these composites compare to current commercial materials. From this characterization and optimization, a natural fiber sample with an R/in. >5 was achieved, making these materials competitive with commercial alternatives.

隔热材料可以显著减少维持建筑物内部温度所需的能源，节省资金并减少对环境的影响。许多工业上可用的绝缘材料都使用石油衍生材料，如聚氨酯、玻璃纤维和聚苯乙烯，这些材料在生产过程中会排放温室气体，对环境有害。这些合成绝缘材料的可持续替代品是使用可再生原料来生产类似的绝缘产品。在这项工作中，亚麻和香蕉等天然纤维与环氧树脂体系结合，以产生天然纤维绝缘。本研究的关键实验参数包括评价纤维长度和加工方法，同时改变环氧粘合剂溶剂溶解的粘度。通过比较密度、导热系数、回弹性和抗压强度来确定保温性能；对这些参数进行了测试，以进一步了解和研究系统内的结构-性能关系，并了解这些复合材料与当前商用材料的比较。从这个表征和优化，天然纤维样品具有R/in。>；5已经实现，使这些材料与商业替代品具有竞争力。

{"title":"Structure-property relationships and recyclability of natural fiber insulation","authors":"Meghan E. Lamm , Bayleigh M. Loving , Katie Copenhaver , Amber M. Hubbard , Caitlyn M. Clarkson","doi":"10.1016/j.buildenv.2025.114156","DOIUrl":"10.1016/j.buildenv.2025.114156","url":null,"abstract":"<div><div>Thermal insulation can significantly decrease the energy required to maintain internal temperatures within buildings, saving money and decreasing environmental impacts. Many industrially available insulation options utilize petroleum-derived materials, such as polyurethane, fiberglass, and polystyrene, which can be detrimental to the environment by emitting greenhouse gases during production. A sustainable alternative to these synthetic insulations is the use of renewable feedstocks to produce comparable insulation products. In this work, natural fibers such as flax and banana were bound with an epoxy system to produce natural-fiber insulation. Key experimental parameters for this study included evaluating the fiber lengths and processing methods while varying the viscosity of the epoxy binder solvent dissolution. Insulation performance was determined by comparing density, thermal conductivity, resilience, and compressive strength; these parameters were tested to further understand and investigate the structure- property relationships within the system and to understand how these composites compare to current commercial materials. From this characterization and optimization, a natural fiber sample with an R/in. >5 was achieved, making these materials competitive with commercial alternatives.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114156"},"PeriodicalIF":7.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838474","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Environmental sounds on office worker performance: Non-invasive monitoring of physiological responses in laboratory settings 环境声音对办公室工作人员表现的影响：实验室环境中生理反应的非侵入性监测

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-19 DOI: 10.1016/j.buildenv.2025.114137

Lei Wang , Jie Ma , Heng Li , Michael C.H. Yam , Zeyu Zhao , Changyuan Yu

Environmental sounds substantially influence occupational health and productivity, particularly in office settings where prolonged background noise exposure coincides with cognitive demands. This study introduces a non-invasive monitoring approach to examine the effects of sound type, intensity, and task difficulty on office workers. A smart cushion was developed to monitor heartbeat and respiration signals, while salivary samples were analyzed for cortisol and glucose. A two-stage controlled experimental protocol was employed. In Stage One, ten participants performed a 2-back working memory task under six standardized sound conditions at 60 dB (LAeq). Based on resulting physiological and performance patterns, these sounds were categorized as positive, neutral, or negative. In Stage Two, thirty new participants were randomly assigned to one of the three soundscape categories and completed four working memory tasks (1-back and 3-back) under two sound intensity levels (40 dB (LAeq) and 80 dB (LAeq)). The results revealed that (1) Different sound types significantly influenced office workers’ comfort perception, task performance, and physiological responses. (2) Higher task difficulty and sound intensity both induced more pronounced physiological responses and reduced performance and comfort. (3) Autonomic measures and cortisol were highly responsive to sound conditions, whereas glucose showed limited sensitivity. This study advances understanding of how environmental sounds shape human health and cognitive functioning in office settings. Evidence from physiological responses provides a robust basis for developing sound management strategies that enhance comfort, well-being, and productivity. These findings support the design of more sustainable and human-centered urban workspaces.

环境噪音严重影响职业健康和生产力，特别是在办公室环境中，长时间接触背景噪音与认知需求相吻合。本研究引入一种非侵入性的监测方法来检验声音类型、强度和任务难度对办公室工作人员的影响。研究人员开发了一种智能垫子来监测心跳和呼吸信号，同时分析唾液样本中的皮质醇和葡萄糖。采用两阶段对照实验方案。在第一阶段，10名被试在6种标准的60分贝（LAeq）的声音条件下进行双背工作记忆任务。根据由此产生的生理和表现模式，这些声音被分为积极的、中性的和消极的。在第二阶段，30名新参与者被随机分配到三个音景类别中的一个，并在两个声强水平（40 dB （LAeq）和80 dB (LAeq)）下完成4个工作记忆任务（1-back和3-back）。结果表明：(1)不同声音类型显著影响办公室员工的舒适感知、任务绩效和生理反应。(2)较高的任务难度和声音强度会引起更明显的生理反应，并降低性能和舒适度。(3)自主神经测量和皮质醇对声音条件有高度反应，而葡萄糖则表现出有限的敏感性。这项研究促进了对环境声音如何在办公室环境中塑造人类健康和认知功能的理解。来自生理反应的证据为制定健全的管理策略提供了坚实的基础，从而提高舒适度、幸福感和生产力。这些发现为设计更加可持续和以人为本的城市工作空间提供了支持。

{"title":"Environmental sounds on office worker performance: Non-invasive monitoring of physiological responses in laboratory settings","authors":"Lei Wang , Jie Ma , Heng Li , Michael C.H. Yam , Zeyu Zhao , Changyuan Yu","doi":"10.1016/j.buildenv.2025.114137","DOIUrl":"10.1016/j.buildenv.2025.114137","url":null,"abstract":"<div><div>Environmental sounds substantially influence occupational health and productivity, particularly in office settings where prolonged background noise exposure coincides with cognitive demands. This study introduces a non-invasive monitoring approach to examine the effects of sound type, intensity, and task difficulty on office workers. A smart cushion was developed to monitor heartbeat and respiration signals, while salivary samples were analyzed for cortisol and glucose. A two-stage controlled experimental protocol was employed. In Stage One, ten participants performed a 2-back working memory task under six standardized sound conditions at 60 dB (LAeq). Based on resulting physiological and performance patterns, these sounds were categorized as positive, neutral, or negative. In Stage Two, thirty new participants were randomly assigned to one of the three soundscape categories and completed four working memory tasks (1-back and 3-back) under two sound intensity levels (40 dB (LAeq) and 80 dB (LAeq)). The results revealed that (1) Different sound types significantly influenced office workers’ comfort perception, task performance, and physiological responses. (2) Higher task difficulty and sound intensity both induced more pronounced physiological responses and reduced performance and comfort. (3) Autonomic measures and cortisol were highly responsive to sound conditions, whereas glucose showed limited sensitivity. This study advances understanding of how environmental sounds shape human health and cognitive functioning in office settings. Evidence from physiological responses provides a robust basis for developing sound management strategies that enhance comfort, well-being, and productivity. These findings support the design of more sustainable and human-centered urban workspaces.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114137"},"PeriodicalIF":7.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145799550","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Comparison of two adjoint-based inverse design methods in the shape optimization of local exhaust hoods for flow resistance reduction 两种基于伴随逆设计方法在局部排气罩减阻形状优化中的比较

IF 7.6 1区工程技术 Q1 CONSTRUCTION & BUILDING TECHNOLOGY

Building and Environment

Pub Date : 2025-12-19 DOI: 10.1016/j.buildenv.2025.114157

Fei Liu , Yinglei Zhang , Yong Chen , Siyi Luo , Zongliang Zuo , Haohan Sha , Tongxin Zhu , Wei Liu

The resistance of local ventilation components leads to considerable fan power consumption, and reducing local resistance is therefore essential for energy savings. Generally, the resistance reduction design can be achieved through shape optimization, which mainly relies on two design variables based on the adjoint method: the cell coordinates that define the geometric boundaries, and the cell porosities within the local component. This study was to compare the two design variables with an emphasis on their performance in shaping a local exhaust hood. The comparison indicated that different locally optimized geometries can be obtained by both design variables, with maximum resistance reduction rates of 76.4% and 50.9% achieved by the design variable of the cell coordinates and porosities, respectively. The shape obtained through porosity-based optimization can be further refined by cell coordinates. The numerical parameters of the adjoint method investigated in this study influenced either the optimized shape or the optimization efficiency when the design variable of cell coordinates was employed. In contrast, for the porosity-based design, the optimized shape was insensitive to numerical parameters, and it could achieve higher computational efficiency without affecting the optimization results. These findings clarified the performance differences between the two adjoint-based design variables and provided guidance for selecting appropriate optimization strategies for local ventilation components.

局部通风部件的阻力导致风机功耗相当大，因此降低局部阻力对节能至关重要。通常通过形状优化来实现减阻设计，形状优化主要依赖于基于伴随方法的两个设计变量：定义几何边界的单元坐标和局部分量内的单元孔隙率。这项研究是比较两个设计变量，重点是它们在塑造局部排气罩方面的表现。结果表明，两种设计变量均可获得不同的局部优化几何形状，以单元坐标和孔隙率为设计变量，最大阻力降低率分别为76.4%和50.9%。通过孔隙度优化得到的形状可以通过单元坐标进一步细化。当采用单元坐标为设计变量时，所研究的伴随方法的数值参数对优化形状和优化效率都有影响。相比之下，基于孔隙度的优化形状对数值参数不敏感，在不影响优化结果的情况下可以实现更高的计算效率。这些发现阐明了两种基于伴随的设计变量之间的性能差异，并为选择适当的局部通风组件优化策略提供了指导。

{"title":"Comparison of two adjoint-based inverse design methods in the shape optimization of local exhaust hoods for flow resistance reduction","authors":"Fei Liu , Yinglei Zhang , Yong Chen , Siyi Luo , Zongliang Zuo , Haohan Sha , Tongxin Zhu , Wei Liu","doi":"10.1016/j.buildenv.2025.114157","DOIUrl":"10.1016/j.buildenv.2025.114157","url":null,"abstract":"<div><div>The resistance of local ventilation components leads to considerable fan power consumption, and reducing local resistance is therefore essential for energy savings. Generally, the resistance reduction design can be achieved through shape optimization, which mainly relies on two design variables based on the adjoint method: the cell coordinates that define the geometric boundaries, and the cell porosities within the local component. This study was to compare the two design variables with an emphasis on their performance in shaping a local exhaust hood. The comparison indicated that different locally optimized geometries can be obtained by both design variables, with maximum resistance reduction rates of 76.4% and 50.9% achieved by the design variable of the cell coordinates and porosities, respectively. The shape obtained through porosity-based optimization can be further refined by cell coordinates. The numerical parameters of the adjoint method investigated in this study influenced either the optimized shape or the optimization efficiency when the design variable of cell coordinates was employed. In contrast, for the porosity-based design, the optimized shape was insensitive to numerical parameters, and it could achieve higher computational efficiency without affecting the optimization results. These findings clarified the performance differences between the two adjoint-based design variables and provided guidance for selecting appropriate optimization strategies for local ventilation components.</div></div>","PeriodicalId":9273,"journal":{"name":"Building and Environment","volume":"290 ","pages":"Article 114157"},"PeriodicalIF":7.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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