東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Knowledge Integrated Deep Learning f...

Su, Yunsheng,

Linked to FindBook

Google Book

Amazon

博客來

Knowledge Integrated Deep Learning for Enhanced Fault Diagnosis and Prognosis for Rotating Machinery and Its Applications on Marine Systems /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Knowledge Integrated Deep Learning for Enhanced Fault Diagnosis and Prognosis for Rotating Machinery and Its Applications on Marine Systems // Yunsheng Su.
Author:	Su, Yunsheng,
Description:	1 electronic resource (149 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 86-11, Section: B.
Contained By:	Dissertations Abstracts International86-11B.
Subject:	Mechanical engineering. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31937962
ISBN:	9798314865156

Knowledge Integrated Deep Learning for Enhanced Fault Diagnosis and Prognosis for Rotating Machinery and Its Applications on Marine Systems /
Su, Yunsheng,

Knowledge Integrated Deep Learning for Enhanced Fault Diagnosis and Prognosis for Rotating Machinery and Its Applications on Marine Systems /Yunsheng Su. - 1 electronic resource (149 pages)

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

This dissertation presents a knowledge integrated deep learning framework for robust fault diagnosis and prognosis, applied across diverse engineered systems, including bearings, lithium-ion batteries, and mooring systems in wave energy converters (WECs). In the context of bearing fault detection, the research develops a deep transfer learning (DTL) approach to fuse data from multiple sources and improve diagnosis accuracy by using Convolutional Neural Networks (CNNs) to extract critical features. While the DTL method enhances fault detection, the focus is on the primary contribution: a knowledge-informed deep network (KIDN) approach.For bearing fault diagnosis, the KIDN approach begins by extracting knowledge-based features derived from engineering principles, which are then used to train a deep network for more accurate fault detection. To further enhance the generalizability and reduce the need for extensive tuning, a generalizability-based adaptive network design strategy is introduced, utilizing a constrained Gaussian process (CGP) model to optimize network architectures. The proposed KIDN framework improves the robustness of fault diagnosis and reduces the computational effort typically required in tuning deep networks. This approach is validated through multiple experimental case studies, demonstrating its efficacy in accurately predicting fault types and severity in rolling bearings.For lithium-ion batteries, a knowledge-constrained machine learning (KcML) framework is developed to predict the remaining useful life (RUL) by incorporating prior knowledge of battery degradation. The knowledge is used as additional constraints to guide the training of machine learning models, improving the accuracy and reliability of RUL predictions in varying operational conditions.The dissertation extends the knowledge-informed approach to the fault diagnosis of WEC mooring systems, where a combination of Autoregressive (AR) models and CNNs is used to detect faults like corrosion and biofouling. The framework effectively classifies fault severity based on dynamic responses such as surge, heave, and pitch motions.By applying knowledge-informed deep learning techniques across these domains, this work advances fault diagnosis and prognostics methods, enhancing system reliability and contributing to the efficient maintenance of industrial and renewable energy systems.

English

ISBN: 9798314865156Subjects--Topical Terms:

649730
Mechanical engineering.
Subjects--Index Terms:

Bearing

Knowledge Integrated Deep Learning for Enhanced Fault Diagnosis and Prognosis for Rotating Machinery and Its Applications on Marine Systems /
LDR:03974nmm a22004453i 4500 001 2434851
005 20251001062539.5
006 m o d
007 cr|nu||||||||
008 261011s2025 miu||||||m |||||||eng d
020 $a 9798314865156
035 $a (MiAaPQD)AAI31937962
035 $a AAI31937962
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Su, Yunsheng, $e author. $0 (orcid)0009-0001-4185-2133 $3 3824978
245 1 0 $a Knowledge Integrated Deep Learning for Enhanced Fault Diagnosis and Prognosis for Rotating Machinery and Its Applications on Marine Systems / $c Yunsheng Su.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2025
300 $a 1 electronic resource (149 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: Zou, Shangyan; Blough, Jason R. Committee members: Wang, Zequn; Sain, Trisha; Cai, Chunpei; Lei, Xinyu.
502 $b Ph.D. $c Michigan Technological University $d 2025.
520 $a This dissertation presents a knowledge integrated deep learning framework for robust fault diagnosis and prognosis, applied across diverse engineered systems, including bearings, lithium-ion batteries, and mooring systems in wave energy converters (WECs). In the context of bearing fault detection, the research develops a deep transfer learning (DTL) approach to fuse data from multiple sources and improve diagnosis accuracy by using Convolutional Neural Networks (CNNs) to extract critical features. While the DTL method enhances fault detection, the focus is on the primary contribution: a knowledge-informed deep network (KIDN) approach.For bearing fault diagnosis, the KIDN approach begins by extracting knowledge-based features derived from engineering principles, which are then used to train a deep network for more accurate fault detection. To further enhance the generalizability and reduce the need for extensive tuning, a generalizability-based adaptive network design strategy is introduced, utilizing a constrained Gaussian process (CGP) model to optimize network architectures. The proposed KIDN framework improves the robustness of fault diagnosis and reduces the computational effort typically required in tuning deep networks. This approach is validated through multiple experimental case studies, demonstrating its efficacy in accurately predicting fault types and severity in rolling bearings.For lithium-ion batteries, a knowledge-constrained machine learning (KcML) framework is developed to predict the remaining useful life (RUL) by incorporating prior knowledge of battery degradation. The knowledge is used as additional constraints to guide the training of machine learning models, improving the accuracy and reliability of RUL predictions in varying operational conditions.The dissertation extends the knowledge-informed approach to the fault diagnosis of WEC mooring systems, where a combination of Autoregressive (AR) models and CNNs is used to detect faults like corrosion and biofouling. The framework effectively classifies fault severity based on dynamic responses such as surge, heave, and pitch motions.By applying knowledge-informed deep learning techniques across these domains, this work advances fault diagnosis and prognostics methods, enhancing system reliability and contributing to the efficient maintenance of industrial and renewable energy systems.
546 $a English
590 $a School code: 0129
650 4 $a Mechanical engineering. $3 649730
650 4 $a Engineering. $3 586835
650 4 $a Alternative energy. $3 3436775
653 $a Bearing
653 $a Fault diagnosis
653 $a Knowledge-informed
653 $a Prognosis
653 $a Deep transfer learning
690 $a 0548
690 $a 0537
690 $a 0363
710 2 $a Michigan Technological University. $b Mechanical and Aerospace Engineering. $e degree granting institution. $3 3817264
720 1 $a Zou, Shangyan $e degree supervisor.
720 1 $a Blough, Jason R. $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-11B.
790 $a 0129
791 $a Ph.D.
792 $a 2025
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31937962