東華大學圖書館 |

Integrating Physical and Data-Driven Perspectives on Building Energy Performance: A Tale of Two Cities.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Integrating Physical and Data-Driven Perspectives on Building Energy Performance: A Tale of Two Cities./
作者:	Nutkiewicz, Alexandra Ilana.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2021,
面頁冊數:	184 p.
附註:	Source: Dissertations Abstracts International, Volume: 83-05, Section: B.
Contained By:	Dissertations Abstracts International83-05B.
標題:	Heat. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28812936
ISBN:	9798494453259

Integrating Physical and Data-Driven Perspectives on Building Energy Performance: A Tale of Two Cities.
Nutkiewicz, Alexandra Ilana.

Integrating Physical and Data-Driven Perspectives on Building Energy Performance: A Tale of Two Cities. - Ann Arbor : ProQuest Dissertations & Theses, 2021 - 184 p.

Source: Dissertations Abstracts International, Volume: 83-05, Section: B.

Thesis (Ph.D.)--Stanford University, 2021.

This item must not be sold to any third party vendors.

Energy is the keystone supporting our everyday lives. As it also the primary driver for the world's ongoing climate crisis, the need to decarbonize the urban built environment is becoming increasingly dire. It is well-established that with a better understanding of how and when buildings consume energy, "low-hanging" solutions in the form of building retrofits can provide key energy, environmental, and economic improvements to cities. It is also critical, given the already warming climate, we utilize intelligent building design to slow the increasing demand for energy. Emerging sensing and data infrastructure, when combined with recent advancements in computational modeling methods, represent a key opportunity to improve the energy sustainability of our cities. Therefore, this dissertation proposes new strategies that leverage emerging data sources to empower urban sustainability stakeholders with information needed to make informed decisions regarding their energy systems and built environment. Rapid growth in sensing technologies has led to a windfall of structured and unstructured data streams describing the urban built environment. Smart meter infrastructure, for example, can provide us with an hourly (or sub-hourly) stream of data describing whole-building energy consumption, versus the monthly or annual frequency we have already had. And with local governments becoming increasingly aware of their roles in achieving the targets of the Paris Climate Agreement and Sustainable Development Goals, open data initiatives are making building, utility, and transportation information newly available to the public. Despite this broad availability of information describing our cities, we lack tools capable of making sense of this data. But when combined with interpretable visualization and computational techniques, they can produce once hidden insights to improve urban building energy performance. While many high-income cities globally are beginning to understand the significance of data on urban energy eciency, low and middle-income cities instead deal with a lack of data availability. Rather than having an abundance of data to support computationally-informed decision-making, development policy in these cities must instead rely on the limited information created through slow, expensive surveys and other administrative initiatives. A key research goal in this dissertation is to demonstrate how varying levels of data can be utilized to improve urban building sustainability. To do so, this work develops computational tools for two types of built environments: data-rich and data-sparse cities. The first context area of study - data-rich cities - explores the capability of using high-fidelity data streams to predict building energy consumption while considering the impacts of the surrounding urban context. Urban context - the built structures and natural environment surrounding a building - is minimally considered in the often-disparate physics-based energy simulation and datadriven machine learning prediction methods. This dissertation introduces a Data-driven Urban Energy Simulation (DUE-S) framework that integrates simulation and deep learning models to better predict the complex spatiotemporal nature of building energy consumption. I demonstrate how the proposed DUE-S model can predict the impacts of the urban context and various large-scale retrofit programs on building energy performance on multiple spatial and temporal scales. I show how high-fidelity datasets and an interpretable modeling framework can empower policymakers to make insightful urban energy planning decisions. The second context area of study focuses on data-sparse cities, specifically informal settlements, or "slums," where little to no data is available to describe the urban built environment. These cities, typically located in low and middle-income countries in the Global South, represent the many challenges associated with climate change including rapid population growth and increased vulnerability to heat-related discomfort. It is also expected that the majority of the future urban building stock in these cities have yet to be constructed. Thus, there exists an opportunity for datadriven design and planning to have a significant impact on the sustainability and energy future of these cities. I introduce a modeling framework that relies on limited sensor data and observational information to evaluate how building design decisions influence the onset of heat stress and demand for energy-intensive space cooling. I show how, despite the limited availability of data describing the built environment, computational tools can still be utilized to inform future urban sustainability policymaking. Overall, this dissertation proposes new data-driven and simulation-based strategies to inform decision makers with information needed to improve urban building energy performance in cities across the world. Despite the unique energy challenges facing every city, creating interpretable, data-driven solutions can inform the future planning of the built environment to improve energy eciency and human well-being. The results and methods I introduce in the following chapters contribute theoretical and practical knowledge to decision makers determining the energy sustainability future of our world.

ISBN: 9798494453259Subjects--Topical Terms:

573595
Heat.

Integrating Physical and Data-Driven Perspectives on Building Energy Performance: A Tale of Two Cities.
LDR:06463nmm a2200337 4500 001 2350324
005 20221020125240.5
008 241004s2021 ||||||||||||||||| ||eng d
020 $a 9798494453259
035 $a (MiAaPQ)AAI28812936
035 $a (MiAaPQ)STANFORDnm585pw3295
035 $a AAI28812936
040 $a MiAaPQ $c MiAaPQ
100 1 $a Nutkiewicz, Alexandra Ilana. $3 3689797
245 1 0 $a Integrating Physical and Data-Driven Perspectives on Building Energy Performance: A Tale of Two Cities.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2021
300 $a 184 p.
500 $a Source: Dissertations Abstracts International, Volume: 83-05, Section: B.
500 $a Advisor: Jain, Rishee; Fischer, Martin A; Gorle, Catherine.
502 $a Thesis (Ph.D.)--Stanford University, 2021.
506 $a This item must not be sold to any third party vendors.
520 $a Energy is the keystone supporting our everyday lives. As it also the primary driver for the world's ongoing climate crisis, the need to decarbonize the urban built environment is becoming increasingly dire. It is well-established that with a better understanding of how and when buildings consume energy, "low-hanging" solutions in the form of building retrofits can provide key energy, environmental, and economic improvements to cities. It is also critical, given the already warming climate, we utilize intelligent building design to slow the increasing demand for energy. Emerging sensing and data infrastructure, when combined with recent advancements in computational modeling methods, represent a key opportunity to improve the energy sustainability of our cities. Therefore, this dissertation proposes new strategies that leverage emerging data sources to empower urban sustainability stakeholders with information needed to make informed decisions regarding their energy systems and built environment. Rapid growth in sensing technologies has led to a windfall of structured and unstructured data streams describing the urban built environment. Smart meter infrastructure, for example, can provide us with an hourly (or sub-hourly) stream of data describing whole-building energy consumption, versus the monthly or annual frequency we have already had. And with local governments becoming increasingly aware of their roles in achieving the targets of the Paris Climate Agreement and Sustainable Development Goals, open data initiatives are making building, utility, and transportation information newly available to the public. Despite this broad availability of information describing our cities, we lack tools capable of making sense of this data. But when combined with interpretable visualization and computational techniques, they can produce once hidden insights to improve urban building energy performance. While many high-income cities globally are beginning to understand the significance of data on urban energy eciency, low and middle-income cities instead deal with a lack of data availability. Rather than having an abundance of data to support computationally-informed decision-making, development policy in these cities must instead rely on the limited information created through slow, expensive surveys and other administrative initiatives. A key research goal in this dissertation is to demonstrate how varying levels of data can be utilized to improve urban building sustainability. To do so, this work develops computational tools for two types of built environments: data-rich and data-sparse cities. The first context area of study - data-rich cities - explores the capability of using high-fidelity data streams to predict building energy consumption while considering the impacts of the surrounding urban context. Urban context - the built structures and natural environment surrounding a building - is minimally considered in the often-disparate physics-based energy simulation and datadriven machine learning prediction methods. This dissertation introduces a Data-driven Urban Energy Simulation (DUE-S) framework that integrates simulation and deep learning models to better predict the complex spatiotemporal nature of building energy consumption. I demonstrate how the proposed DUE-S model can predict the impacts of the urban context and various large-scale retrofit programs on building energy performance on multiple spatial and temporal scales. I show how high-fidelity datasets and an interpretable modeling framework can empower policymakers to make insightful urban energy planning decisions. The second context area of study focuses on data-sparse cities, specifically informal settlements, or "slums," where little to no data is available to describe the urban built environment. These cities, typically located in low and middle-income countries in the Global South, represent the many challenges associated with climate change including rapid population growth and increased vulnerability to heat-related discomfort. It is also expected that the majority of the future urban building stock in these cities have yet to be constructed. Thus, there exists an opportunity for datadriven design and planning to have a significant impact on the sustainability and energy future of these cities. I introduce a modeling framework that relies on limited sensor data and observational information to evaluate how building design decisions influence the onset of heat stress and demand for energy-intensive space cooling. I show how, despite the limited availability of data describing the built environment, computational tools can still be utilized to inform future urban sustainability policymaking. Overall, this dissertation proposes new data-driven and simulation-based strategies to inform decision makers with information needed to improve urban building energy performance in cities across the world. Despite the unique energy challenges facing every city, creating interpretable, data-driven solutions can inform the future planning of the built environment to improve energy eciency and human well-being. The results and methods I introduce in the following chapters contribute theoretical and practical knowledge to decision makers determining the energy sustainability future of our world.
590 $a School code: 0212.
650 4 $a Heat. $3 573595
650 4 $a Deep learning. $3 3554982
650 4 $a Energy modeling. $3 3683859
650 4 $a Energy consumption. $3 631630
650 4 $a Morphology. $3 591167
650 4 $a Artificial intelligence. $3 516317
650 4 $a Biology. $3 522710
650 4 $a Energy. $3 876794
690 $a 0287
690 $a 0800
690 $a 0306
690 $a 0791
710 2 $a Stanford University. $3 754827
773 0 $t Dissertations Abstracts International $g 83-05B.
790 $a 0212
791 $a Ph.D.
792 $a 2021
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28812936