摘要
基于PCA和K-means聚类的算法,采用LabVIEW编写了程序界面交互友好的操作软件系统,用于研究纸张在不同变量上的相似程度。程序设计主要赋予了简便操作,人机友好,数据可视化的特点。基于不同的多变量,采用PCA结合K-means聚类方法对纸张进行了相似程度的分析研究,结果表明:程序界面具有清晰简洁、交互友好、可操作性强的特点,可对纸张进行聚类分析,且可信度高。本程序可用于不同品质等级纸张在不同多变量上差异的识别和相似品质纸张的聚类,有效地进行纸张的各种差异化的分析。
Based on the PCA and K-means clustering algorithm, LabVIEW was used to write a program-friendly interactive operating software system for studying the similarity of paper with different variables. The program design mainly gives the characteristics of easy operation, human-machine friendly and data visualization. Based on different multivariates, PCA combined with K-means clustering method was used to analyze the similarity degree of paper. The results show that the program interface is clear and concise, interactive and operability, and can be used to clustering analyze paper and has high credibility. This program can be used for the recognition of differences in different multi-variable papers of different quality grades and the clustering of similar quality papers, effectively analyzing various kinds of papers.
作者
何智恒
戴毅
柴欣生
HE Zhiheng;DAI Yi;CHAI Xinsheng(State Key Laboratory of Pulp and Paper Engineering, South China University of Technology,Guangzhou 510640, China;National Paper Product Quality Supervision Inspection Center, Dongguan 523080,China)
出处
《造纸科学与技术》
2019年第3期14-23,共10页
Paper Science & Technology
基金
国家自然科学基金项目(21576105)
