Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection

Muhammad Hussain; Tianhua Chen; Sofya Titarenko; Richard Hill; Hussain Al-Aqrabi

doi:10.1007/978-3-031-55568-8_35

Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection

Muhammad Hussain, Tianhua Chen, Sofya Titarenko, Richard Hill, Hussain Al-Aqrabi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This paper presents a framework for the detection of cracked Photovoltaic (PV) cells within the production environment through the creation of a Convolutional Neural Network (CNN). The paper demonstrates how a simple CNN using certain data augmentation techniques and specific regularization i.e., batch normalisation is efficient for PV based crack detection, achieving a recall rate of 99.2% and F1-score of 97.4%. We validate each methodological iteration via KFold cross validation providing granular metrical details allowing us to further optimize the model recall metric to suit the needs of our end deployment environment. We understand and appreciate the lack of highly quality PV image data of defect and therefore through our research we propose specific data augmentations for scaling and injecting variance into PV datasets. The augmentations do not provide any artificial propping of the model performance but are rather cases that may be found on the production line for PV cell manufacturing.

Original language	English
Title of host publication	Advances in Computational Intelligence Systems
Subtitle of host publication	Contributions Presented at the 21st UK Workshop on Computational Intelligence, September 7-9, 2022, Sheffield, UK
Editors	George Panoutsos, Mahdi Mahfouf, Lyudmila S. Mihaylova
Publisher	Springer, Cham
Pages	415-426
Number of pages	12
Edition	1st
ISBN (Electronic)	9783031555688
ISBN (Print)	9783031555671
DOIs	https://doi.org/10.1007/978-3-031-55568-8_35
Publication status	Published - 19 May 2024
Event	21st UK Workshop on Computational Intelligence - University of Sheffield, Sheffield, United Kingdom Duration: 7 Sep 2022 → 9 Sep 2022 Conference number: 21 https://www.sheffield.ac.uk/ukci2022

Publication series

Name	Advances in Intelligent Systems and Computing
Publisher	Springer Cham
ISSN (Print)	2194-5357
ISSN (Electronic)	2194-5365

Workshop

Workshop	21st UK Workshop on Computational Intelligence
Abbreviated title	UKCI 2022
Country/Territory	United Kingdom
City	Sheffield
Period	7/09/22 → 9/09/22
Internet address	https://www.sheffield.ac.uk/ukci2022

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-031-55568-8_35Licence: Unspecified

Cite this

Hussain, M., Chen, T., Titarenko, S., Hill, R., & Al-Aqrabi, H. (2024). Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection. In G. Panoutsos, M. Mahfouf, & L. S. Mihaylova (Eds.), Advances in Computational Intelligence Systems: Contributions Presented at the 21st UK Workshop on Computational Intelligence, September 7-9, 2022, Sheffield, UK (1st ed., pp. 415-426). (Advances in Intelligent Systems and Computing). Springer, Cham. https://doi.org/10.1007/978-3-031-55568-8_35

Hussain, Muhammad ; Chen, Tianhua ; Titarenko, Sofya et al. / Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection. Advances in Computational Intelligence Systems: Contributions Presented at the 21st UK Workshop on Computational Intelligence, September 7-9, 2022, Sheffield, UK. editor / George Panoutsos ; Mahdi Mahfouf ; Lyudmila S. Mihaylova. 1st. ed. Springer, Cham, 2024. pp. 415-426 (Advances in Intelligent Systems and Computing).

@inproceedings{946896b396474cc3b9a7427889ac136c,

title = "Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection",

abstract = "This paper presents a framework for the detection of cracked Photovoltaic (PV) cells within the production environment through the creation of a Convolutional Neural Network (CNN). The paper demonstrates how a simple CNN using certain data augmentation techniques and specific regularization i.e., batch normalisation is efficient for PV based crack detection, achieving a recall rate of 99.2% and F1-score of 97.4%. We validate each methodological iteration via KFold cross validation providing granular metrical details allowing us to further optimize the model recall metric to suit the needs of our end deployment environment. We understand and appreciate the lack of highly quality PV image data of defect and therefore through our research we propose specific data augmentations for scaling and injecting variance into PV datasets. The augmentations do not provide any artificial propping of the model performance but are rather cases that may be found on the production line for PV cell manufacturing. ",

keywords = "Photovoltaic, Deep learning, Convolutional Networks, Batch Normalisation, Data Augmentation",

author = "Muhammad Hussain and Tianhua Chen and Sofya Titarenko and Richard Hill and Hussain Al-Aqrabi",

year = "2024",

month = may,

day = "19",

doi = "10.1007/978-3-031-55568-8_35",

language = "English",

isbn = "9783031555671",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer, Cham",

pages = "415--426",

editor = "George Panoutsos and Mahdi Mahfouf and Mihaylova, {Lyudmila S.}",

booktitle = "Advances in Computational Intelligence Systems",

address = "Switzerland",

edition = "1st",

note = "21st UK Workshop on Computational Intelligence, UKCI 2022 ; Conference date: 07-09-2022 Through 09-09-2022",

url = "https://www.sheffield.ac.uk/ukci2022",

}

Hussain, M , Chen, T, Titarenko, S, Hill, R & Al-Aqrabi, H 2024, Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection. in G Panoutsos, M Mahfouf & LS Mihaylova (eds), Advances in Computational Intelligence Systems: Contributions Presented at the 21st UK Workshop on Computational Intelligence, September 7-9, 2022, Sheffield, UK. 1st edn, Advances in Intelligent Systems and Computing, Springer, Cham, pp. 415-426, 21st UK Workshop on Computational Intelligence, Sheffield, United Kingdom, 7/09/22. https://doi.org/10.1007/978-3-031-55568-8_35

Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection. / Hussain, Muhammad ; Chen, Tianhua; Titarenko, Sofya et al.

Advances in Computational Intelligence Systems: Contributions Presented at the 21st UK Workshop on Computational Intelligence, September 7-9, 2022, Sheffield, UK. ed. / George Panoutsos; Mahdi Mahfouf; Lyudmila S. Mihaylova. 1st. ed. Springer, Cham, 2024. p. 415-426 (Advances in Intelligent Systems and Computing).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection

AU - Hussain, Muhammad

AU - Chen, Tianhua

AU - Titarenko, Sofya

AU - Hill, Richard

AU - Al-Aqrabi, Hussain

N1 - Conference code: 21

PY - 2024/5/19

Y1 - 2024/5/19

N2 - This paper presents a framework for the detection of cracked Photovoltaic (PV) cells within the production environment through the creation of a Convolutional Neural Network (CNN). The paper demonstrates how a simple CNN using certain data augmentation techniques and specific regularization i.e., batch normalisation is efficient for PV based crack detection, achieving a recall rate of 99.2% and F1-score of 97.4%. We validate each methodological iteration via KFold cross validation providing granular metrical details allowing us to further optimize the model recall metric to suit the needs of our end deployment environment. We understand and appreciate the lack of highly quality PV image data of defect and therefore through our research we propose specific data augmentations for scaling and injecting variance into PV datasets. The augmentations do not provide any artificial propping of the model performance but are rather cases that may be found on the production line for PV cell manufacturing.

AB - This paper presents a framework for the detection of cracked Photovoltaic (PV) cells within the production environment through the creation of a Convolutional Neural Network (CNN). The paper demonstrates how a simple CNN using certain data augmentation techniques and specific regularization i.e., batch normalisation is efficient for PV based crack detection, achieving a recall rate of 99.2% and F1-score of 97.4%. We validate each methodological iteration via KFold cross validation providing granular metrical details allowing us to further optimize the model recall metric to suit the needs of our end deployment environment. We understand and appreciate the lack of highly quality PV image data of defect and therefore through our research we propose specific data augmentations for scaling and injecting variance into PV datasets. The augmentations do not provide any artificial propping of the model performance but are rather cases that may be found on the production line for PV cell manufacturing.

KW - Photovoltaic

KW - Deep learning

KW - Convolutional Networks

KW - Batch Normalisation

KW - Data Augmentation

UR - https://doi.org/10.1007/978-3-031-55568-8

U2 - 10.1007/978-3-031-55568-8_35

DO - 10.1007/978-3-031-55568-8_35

M3 - Conference contribution

SN - 9783031555671

T3 - Advances in Intelligent Systems and Computing

SP - 415

EP - 426

BT - Advances in Computational Intelligence Systems

A2 - Panoutsos, George

A2 - Mahfouf, Mahdi

A2 - Mihaylova, Lyudmila S.

PB - Springer, Cham

T2 - 21st UK Workshop on Computational Intelligence

Y2 - 7 September 2022 through 9 September 2022

ER -

Hussain M , Chen T, Titarenko S, Hill R, Al-Aqrabi H. Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection. In Panoutsos G, Mahfouf M, Mihaylova LS, editors, Advances in Computational Intelligence Systems: Contributions Presented at the 21st UK Workshop on Computational Intelligence, September 7-9, 2022, Sheffield, UK. 1st ed. Springer, Cham. 2024. p. 415-426. (Advances in Intelligent Systems and Computing). Epub 2024 May 18. doi: 10.1007/978-3-031-55568-8_35

Development of A Convolutional Neural Network Architecture for Production Based Photovoltaic Fault Detection

Abstract

Publication series

Workshop

UN SDGs

Access to Document

Other files and links

Fingerprint

PV-CrackNet Architecture for Filter Induced Augmentation and Micro-Cracks Detection within a Photovoltaic Manufacturing Facility