Biomimetic model of photovoltaic cell defect detection based on mimic vision

Zhaoyang Qu, Jiye Zang, Lingcong Li, Yunchang Dong and Nan Qu

Applied Energy, 2024, vol. 376, issue PA, No S0306261924014168

Abstract: Solar energy plays an important role in new power systems, and defect detection on photovoltaic (PV) cells is becoming more and more important for the transition of power systems to clean energy. Traditional target detection models are difficult to identify tiny defects on PV cells with complex textures, for this reason, we propose anthropomorphic vision biomimetic detection models. Firstly, a backbone network inspired by the human sensory field and peripheral vision mechanisms is proposed to design the biomimetic visual attention mechanism as well as the biomimetic feature extraction module to fully extract the dynamic context and perceive the peripheral visual attention, and correlate the two in order to capture the fine-grained features of the defective target in the noise-filled background. Second, in the feature fusion stage, a separated spatial semantic fusion pyramid is designed based on the human brain information transfer mode, and semantic and spatial information transfer modules are designed in different information transfer paths to enhance the ability of defective features to express spatial and semantic information. Then, inspired by the partitioning mechanism of the brain cortex, we propose a cascade detection head for task alignment, adaptive regulatory modulation of features at different scales, alignment of spatial mismatches in classification and localization tasks, and separated design of the structure of classification and localisation branches. Finally, the effectiveness of the model is demonstrated experimentally.

Keywords: Defect detection; Deep learning; Transformer; Photovoltaic cell (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0306261924014168
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:appene:v:376:y:2024:i:pa:s0306261924014168

Ordering information: This journal article can be ordered from
http://www.elsevier.com/wps/find/journaldescription.cws_home/405891/bibliographic
http://www.elsevier. ... 405891/bibliographic

DOI: 10.1016/j.apenergy.2024.124033

Access Statistics for this article

Applied Energy is currently edited by J. Yan

More articles in Applied Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().