Classification of acoustic emission sources produced by carbon/epoxy composite based on support vector machine

2015 Global Conference on Polymer and Composite Materials

材料科学;化学

Ding, Peng^1 ; Li, Qin^1 ; Huang, Xunlei^1

Standard and Quality Control Research Institute, Ministry of Water Resources, Hangzhou

310012, China^1

关键词: Acoustic emission signal;    Acoustic emission sources;    AE signals;    BP neural networks;    Carbon/epoxy;    Carbon/epoxy composites;    Input parameter;    Training sets;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/87/1/012002/pdf
DOI  :  10.1088/1757-899X/87/1/012002

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

Carbon/epoxy specimens were made and stretched to fracture. In the process, acoustic emission (AE) signals were collected and their parameters were set as the input parameters of the neural network. Results show that using support vector machine (SVM) network can recognize the difference of AE sources more accurately than using the BP neural network. In addition, the accuracy of the SVM increases when the number of the training set increases. It is proved that using AE signal parameters and SVM network can recognize the AE sources' pattern well. Published under licence by IOP Publishing Ltd