| 姜红,马星煜.差分拉曼光谱结合PCA-RCSC-Transformer对快递面单的检验研究[J].中国造纸,2025,44(11):172-176 |
二维码（扫一下试试看！） |
| 差分拉曼光谱结合PCA-RCSC-Transformer对快递面单的检验研究 |
| Differential Raman Spectroscopy Combined with PCA-RCSC and Improved Transformer for Courier Face Sheets Inspection Research |
| 收稿日期：2025-04-03 修订日期：2025-06-27 |
| DOI：10.11980/j.issn.0254-508X.2025.11.023 |
| 关键词: 差分拉曼光谱 快递面单 正交约束主成分分析 元素比例-余弦相似度聚类 |
| Key Words:Differential Raman spectroscopy courier face sheets orthogonally constrained principal component analysis elemental ratio-cosine similarity clustering |
| 基金项目:安徽公安学院校级科研项目（2024xjkyyb08）。 |
|
| 摘要点击次数: 1232 |
| 全文下载次数: 555 |
| 摘要:针对热敏纸类快递面单字迹易消退、填料成分稳定的特点，本研究通过差分拉曼光谱对173个不同品牌及打印时间的快递面单样本数据进行采集，提出了一种结合改进正交约束主成分分析（PCA）与元素比例-余弦相似度聚类（RCSC）的方法，并引入稀疏注意力机制的Transformer模型进行数据分类预测。结果表明，利用正交约束PCA对差分拉曼光谱数据进行降维，压缩率达95.6%，结合人工验证的RCSC可将样本分为4类。进一步采用引入稀疏注意力机制的Transformer模型进行分类，预测准确率提升至90.0%，显著优于随机森林、支持向量机等传统方法。 |
| Abstract:Addressing the challenges of easily fading handwriting and stable filler components in thermal paper-based courier face sheets， this study collected data from 173 express delivery label samples from various brands and printing dates through differential Raman spectroscopy， and proposed a novel method integrating modified orthogonally constrained principal component analysis （PCA）， and element ratio-cosine similarity clustering (RCSC)， combined with a Transformer model incorporating a sparse attention mechanism for data classification prediction. The results showed that orthogonally constrained PCA reduced the dimension of differential Raman spectral data and resulted in a compression rate of 95.6%， while RCSC supplemented by manual validation， categorized the samples into four classes. Further classification using the sparse attention-based Transformer model achieved an forecast accuracy of 90.0%， significantly outperforming traditional methods such as random forest and support vector machines. |
| 查看全文
HTML
查看/发表评论 下载PDF阅读器 |