東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁 到查詢結果 [ null ]

切換: 標籤 | MARC模式 | ISBD

Machine Learning Driven Optimization...

Miklaszewski, Jennifer A.,

FindBook

Google Book

Amazon

博客來

Machine Learning Driven Optimization of Complex Turbulent Flows /

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Machine Learning Driven Optimization of Complex Turbulent Flows // Jennifer A Miklaszewski.
作者:	Miklaszewski, Jennifer A.,
面頁冊數:	1 electronic resource (181 pages)
附註:	Source: Dissertations Abstracts International, Volume: 86-11, Section: B.
Contained By:	Dissertations Abstracts International86-11B.
標題:	Computational physics. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31936978
ISBN:	9798314899045

Machine Learning Driven Optimization of Complex Turbulent Flows /
Miklaszewski, Jennifer A.,

Machine Learning Driven Optimization of Complex Turbulent Flows /Jennifer A Miklaszewski. - 1 electronic resource (181 pages)

Source: Dissertations Abstracts International, Volume: 86-11, Section: B.

Optimizing complex turbulent flows presents a difficult challenge due to the non-linear, chaotic nature of turbulence. Most fluid flows found in nature or in engineering applications are turbulent, prompting the need for a design optimization method that can quickly and efficiently handle this complexity. Incorporating machine learning and artificial intelligence can greatly accelerate traditionally computationally-intensive methods of optimizing these flows. The aim of this thesis is to showcase the development of a continuously-learning, machine-learning driven optimization method. This is coupled with computational fluid dynamics (CFD) and is applied to various engineering design problems to demonstrate its potential as a powerful engineering design tool. The first case study presented is the optimization of the exit turbulence field in a gas turbine combustor simulator. Combustor turbulence in a gas turbine engine greatly influences the efficiency of the downstream high pressure turbine stage. Studies have shown that combustor turbulence can result in a 1.3% reduction in stage efficiency of the turbine. This is a staggering number when considering the impact on engine fuel efficiency, with improvements of 0.1% generally garnering substantial research funding. With an optimized design, I am able to achieve substantial improvements in stage efficiency, as well as identify the specific aspects of combustor geometry that contribute to this finding. The second case study presented is the optimization of airflow patterns in a dental office to minimize infection spread. The placement of the air supply and return vents is analyzed to produce optimal circulation within the building and decrease infection potential for dental personnel and patients. Two more cases are presented as tutorial cases for the packaged optimizer and to expand functionality of the optimization scheme, named the Multi-fidelity Integrated Learning Optimization, or MILO. The third case is a simplified heat exchanger, where the cooling air temperature, velocity, and fan speed is adjusted to target a desired outlet water temperature. Uncertainty in the inlet conditions and fan speed is propagated throughout the optimization, showing the potential success of this tool when used in real-world industrial applications where operating conditions may vary. The fourth case introduces multi-objective optimization of a fire suppression system where both the mass flow rate of the water and the amount of solid burned are minimized. This demonstration of a multi-objective problem radically expands the number of engineering design problems that MILO can be applied to. In all cases presented in this thesis, there is a substantial improvement in design performance despite the numerous design constraints applied, as well as an average 88% savings in the number of CPU hours needed to reach an optimized solution.

English

ISBN: 9798314899045Subjects--Topical Terms:

3343998
Computational physics.
Subjects--Index Terms:

Combustor-turbine interaction

Machine Learning Driven Optimization of Complex Turbulent Flows /
LDR:04406nmm a22004333i 4500 001 2432670
005 20250731101909.5
006 m o d
007 cr|nu||||||||
008 261011s2025 miu||||||m |||||||eng d
020 $a 9798314899045
035 $a (MiAaPQD)AAI31936978
035 $a AAI31936978
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Miklaszewski, Jennifer A., $e author. $0 (orcid)0000-0002-8233-0112 $3 3821981
245 1 0 $a Machine Learning Driven Optimization of Complex Turbulent Flows / $c Jennifer A Miklaszewski.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2025
300 $a 1 electronic resource (181 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 86-11, Section: B.
500 $a Advisors: Hamlington, Peter Committee members: Folk, Masha; Henze, Daven; Mukherjee, Debanjan; Miller, Shelly; Evans, John.
502 $b Ph.D. $c University of Colorado at Boulder $d 2025.
520 $a Optimizing complex turbulent flows presents a difficult challenge due to the non-linear, chaotic nature of turbulence. Most fluid flows found in nature or in engineering applications are turbulent, prompting the need for a design optimization method that can quickly and efficiently handle this complexity. Incorporating machine learning and artificial intelligence can greatly accelerate traditionally computationally-intensive methods of optimizing these flows. The aim of this thesis is to showcase the development of a continuously-learning, machine-learning driven optimization method. This is coupled with computational fluid dynamics (CFD) and is applied to various engineering design problems to demonstrate its potential as a powerful engineering design tool. The first case study presented is the optimization of the exit turbulence field in a gas turbine combustor simulator. Combustor turbulence in a gas turbine engine greatly influences the efficiency of the downstream high pressure turbine stage. Studies have shown that combustor turbulence can result in a 1.3% reduction in stage efficiency of the turbine. This is a staggering number when considering the impact on engine fuel efficiency, with improvements of 0.1% generally garnering substantial research funding. With an optimized design, I am able to achieve substantial improvements in stage efficiency, as well as identify the specific aspects of combustor geometry that contribute to this finding. The second case study presented is the optimization of airflow patterns in a dental office to minimize infection spread. The placement of the air supply and return vents is analyzed to produce optimal circulation within the building and decrease infection potential for dental personnel and patients. Two more cases are presented as tutorial cases for the packaged optimizer and to expand functionality of the optimization scheme, named the Multi-fidelity Integrated Learning Optimization, or MILO. The third case is a simplified heat exchanger, where the cooling air temperature, velocity, and fan speed is adjusted to target a desired outlet water temperature. Uncertainty in the inlet conditions and fan speed is propagated throughout the optimization, showing the potential success of this tool when used in real-world industrial applications where operating conditions may vary. The fourth case introduces multi-objective optimization of a fire suppression system where both the mass flow rate of the water and the amount of solid burned are minimized. This demonstration of a multi-objective problem radically expands the number of engineering design problems that MILO can be applied to. In all cases presented in this thesis, there is a substantial improvement in design performance despite the numerous design constraints applied, as well as an average 88% savings in the number of CPU hours needed to reach an optimized solution.
546 $a English
590 $a School code: 0051
650 4 $a Computational physics. $3 3343998
650 4 $a Aerospace engineering. $3 1002622
650 4 $a Fluid mechanics. $3 528155
650 4 $a Mechanical engineering. $3 649730
653 $a Combustor-turbine interaction
653 $a Computational fluid dynamics
653 $a Optimization
653 $a Turbulence
690 $a 0216
690 $a 0538
690 $a 0548
690 $a 0204
710 2 $a University of Colorado at Boulder. $b Mechanical Engineering. $e degree granting institution. $3 3821982
720 1 $a Hamlington, Peter $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-11B.
790 $a 0051
791 $a Ph.D.
792 $a 2025
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31936978