Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Dec 12;25(24):7570.
doi: 10.3390/s25247570.

Enhanced Low-Energy Impact Localization for Carbon-Fiber Honeycomb Sandwich Panels Using LightGBM

Zifan He  1 , Jiyun Lu  1 , Shengming Cui  2 , Chunhua Zhou  3 , Yinuo Shao  4 , Qi Wu  2 , Hongfu Zuo  1
Affiliations

Enhanced Low-Energy Impact Localization for Carbon-Fiber Honeycomb Sandwich Panels Using LightGBM

Zifan He et al. Sensors (Basel). .

Abstract

Low-energy impacts have been demonstrated to cause damage and failure in aircraft structures, thereby affecting the structural load-bearing performance and creating safety hazards. In this study, an innovative damage-monitoring method based on a fiber Bragg grating (FBG) is proposed for honeycomb sandwich composites. The proposed method is applicable to honeycomb sandwich composites and integrates a light gradient boosting machine (LightGBM)-optimized impact localization method with feature-parallel and data-parallel processing in the machine learning architecture. An impact localization algorithm is applied to honeycomb sandwich composites using an array of multiplexed FBG sensors. The proposed algorithm exhibited substantial localization accuracy. The LightGBM method was employed to identify the optimal branching points for impact localization in real time, addressing the low-accuracy challenge in localizing low-energy impacts on the board structure when the fiber grating sensing system operates at a high sampling frequency.

Keywords: fiber Bragg grating sensor; honeycomb sandwich composites; layout optimization; low-energy impact localization; machine learning.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflicts of interest. The funders had no role in the design of this study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

Figures

Figure 1
Figure 1
Flowchart of the low-energy impact localization method using the machine-learning model with multi-domain features for a honeycomb sandwich composite panel. TFn×60 refers to the time-domain features extracted from the shock signal, where “n” indicates the number of samples per window, and “60” refers to the number of time-domain features extracted from a 60-sample window. FFn×20 refers to frequency-domain features derived from the signal’s fast Fourier transform and SFn×60 refers to time–frequency-domain features calculated from the short-time Fourier transform.
Figure 2
Figure 2
Schematic of the test piece. (a) Schematic of the test piece. (b) Specimen.
Figure 3
Figure 3
Sensor layout.
Figure 4
Figure 4
Schematic of the impact loading device.
Figure 5
Figure 5
Experimental setup for the impact loading device.
Figure 6
Figure 6
Measured impact signals from a commercial high-speed FBG interrogator.
Figure 7
Figure 7
Wavelet packet energy decomposition features of the six FBG sensors. (a–f) Feature vectors of FBG1–FBG6, respec-tively. Note: “wp_xxx” corresponds to the energy of a signal in a specific frequency subband. The path name consists of a/d (a = low frequency, d = high frequency); for example, “wp_add” refers to energy in the low–high–high frequency subbands.
Figure 8
Figure 8
Schematic of composite honeycomb panel low-energy impact signal feature processing.
Figure 9
Figure 9
LightGBM model structure and experimental flowchart. In this figure, the symbol “#” indicates “number of”. Specifically, #data represents the number of data samples and #bins represents the number of bins.
Figure 10
Figure 10
Comparison of the accuracy of machine-learning model sensor layout optimization.
Figure 11
Figure 11
Comparison of the predicted and actual locations.
Figure 12
Figure 12
Localization-error analysis of the four regression models.
Figure 13
Figure 13
Low-velocity impact localization results of the honeycomb sandwich composites based on LightBGM modeling.
See this image and copyright information in PMC

References

    1. Gorgin R., Wang Z., Wu Z., Yang Y. Probability based impact localization in plate structures using an error index. Mech. Syst. Signal Process. 2021;157:107724. doi: 10.1016/j.ymssp.2021.107724. - DOI
    1. Seno A.H., Sharif Khodaei Z.S., Aliabadi M.H.F. Passive sensing method for impact localisation in composite plates under simulated environmental and operational conditions. Mech. Syst. Signal Process. 2019;129:20–36. doi: 10.1016/j.ymssp.2019.04.023. - DOI
    1. Aly K., Bradford P.D. Real-time impact damage sensing and localization in composites through embedded aligned carbon nanotube sheets. Compos. B Eng. 2019;162:522–531. doi: 10.1016/j.compositesb.2018.12.104. - DOI
    1. Wang W., Zhang H., Lynch J.P., Cesnik C.E.S., Li H. Experimental and numerical validation of guided wave phased arrays integrated within standard data acquisition systems for structural health monitoring. Struct. Control Health Monit. 2018;25:e2171. doi: 10.1002/stc.2171. - DOI
    1. Su Z., Ye L., Lu Y. Guided Lamb waves for identification of damage in composite structures: A review. J. Sound Vib. 2006;295:753–780. doi: 10.1016/j.jsv.2006.01.020. - DOI

LinkOut - more resources