A PINN-CNN Hybrid Architecture for Vibration-Based Damage Detection in Airship Envelopes

Fangyuan Zhao (), Yiwei Cheng (), Keqiang Xie () and Yuanhang Wang ()
Additional contact information
Fangyuan Zhao: China University of Geosciences (Wuhan), School of Mechanical Engineering and Electronic Information
Yiwei Cheng: China University of Geosciences (Wuhan), School of Mechanical Engineering and Electronic Information
Keqiang Xie: Guangdong Provincial Key Laboratory of Electronic Information Products Reliability Technology
Yuanhang Wang: Shenzhen Technology University, Sino-German College of Intelligent Manufacturing

A chapter in Data-Driven Methods for Reliability and Safety Engineering: Applications in Industrial Systems, 2026, pp 361-373 from Springer

Abstract: Abstract Conventional methods for detecting damage in flexible airship envelopes face limitations in accurately evaluating local material properties under conditions of large deformation. A Physics-Informed Neural Networks Convolutional Neural Networks (PINN-CNN) hybrid architecture is proposed, integrating physical constraints derived from wave equations and membrane dynamics into the neural network training process. The approach was evaluated using synthetic airship envelope data under three conditions: intact, 2 × 2 cross-cut damage, and 5 × 5 cross-cut damage. Experimental findings indicate that the PINN-CNN model attained 100% accuracy, representing a notable improvement over the traditional CNN, which achieved an accuracy of 96.67%, while also exhibiting superior robustness to noise. Under conditions with 25% noise, the PINN-CNN model maintained an accuracy of 90.00%, whereas the standard CNN performance declined to 70.00%.

Keywords: Damage detection; Vibration signal analysis; Noise robustness; Physics-informed neural networks (search for similar items in EconPapers)
Date: 2026
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

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:spr:ssrchp:978-3-032-22873-4_26

Ordering information: This item can be ordered from
http://www.springer.com/9783032228734

DOI: 10.1007/978-3-032-22873-4_26

Access Statistics for this chapter

More chapters in Springer Series in Reliability Engineering from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().