Non-Destructive Identification of Virgin Cashmere and Chemically Modified Wool Fibers Based on Fractional Order Derivative and Improved Wavelength Extraction Algorithm Using NIR Spectroscopy and Chemometrics
Yaolin Zhu Yong Zhang Xin Chen Jinni Chen Hongsong Zhang a School of Electronics and Information,Xi'an Polytechnic University,Xi'an,Chinab Production Technology Department,Shanghai Ranzi Industrial Co. Ltd,Shanghai,China
DOI: https://doi.org/10.1080/15440478.2024.2409901
2024-10-02
Journal of Natural Fibers
Abstract:Virgin cashmere is highly prized for its superior quality, and chemically modified wool is often used to fake it. However, traditional spectral analysis and image processing methods struggle to identify the two fibers. To address this, combining chemometrics with near-infrared (NIR) spectroscopy is proposed to identify virgin cashmere and chemically modified wool based on fractional order derivatives and improved wavelength extraction algorithm. The original spectra of the two fibers are very close, making identification difficult. Therefore, fractional order derivatives are used to amplify tiny spectral differences, which are often overlooked in conventional analysis. Meanwhile, common wavelength extraction algorithms few consider inherent relationship between spectral features and chemical properties, so an improved wavelength extraction algorithm using Shuffled Frog Leaping Algorithm (SFLA) and Beluga Whale Optimization (BWO) is employed to explore the connection of the two. The proposed method with Partial Least Squares Discriminant Analysis (PLS-DA) achieves an accuracy of 94.4%. Experimental results demonstrate that fractional order derivatives enhance tiny spectral features and differences, while extracted feature wavelengths correspond to main chemical characteristic bands of the fibers. This study shows the feasibility of identifying virgin cashmere and chemically modified wool and offers a novel idea for identifying fibers in the textile industry. 原羊绒因其卓越的品质而备受珍视,化学改性羊毛经常被用来伪造它. 然而,传统的光谱分析和图像处理方法难以识别这两种纤维. 为了解决这个问题,提出了将化学计量学与近红外光谱相结合,基于分数阶导数和改进的波长提取算法来识别原羊绒和化学改性羊毛. 这两种纤维的原始光谱非常接近难以识别。因此,分数阶导数被用来放大微小的光谱差异,而这些差异在传统分析中往往被忽视。. 同时,常见的波长提取算法很少考虑光谱特征和化学性质之间的内在关系,因此采用了一种改进的波长提取算法,即使用随机蛙跃算法(SFLA)和白鲸优化算法(BWO)来探索两者之间的联系。. 该方法结合偏最小二乘判别分析(PLS-DA),准确率达到94.4%。. 实验结果表明,分数阶导数增强了微小的光谱特征和差异,而提取的特征波长对应于纤维的主要化学特征带. 本研究表明了识别原羊绒和化学改性羊毛的可行性,并为纺织工业中识别纤维提供了一种新思路.
materials science, textiles