近红外光谱技术的水果成分检测数学模型分析  

Analysis of mathematical model of fruit composition detection by near infrared spectroscopy

作　　者：朱垚 骆利华 ZHU Yao;LUO Lihua(Department of Public Basic Course,Wuhan Institute of Design and Sciences,Wuhan 430000,China;Gant Software ＆ System（Shanghai）Co.Ltd.,Shanghai 200000,China)

机构地区：[1]武汉设计工程学院,武汉430000 [2]甘棠软件系统(上海)有限公司,上海200000

出　　处：《激光杂志》2018年第8期109-112,共4页Laser Journal

基　　金：湖北省高等学校省级教学研究项目(No.2012458)

摘　　要：为了提高水果成分的识别分析能力,针对水果成分检测正确率低的问题,提出基于近红外光谱技术的水果成分检测数学模型。首先采用近红外光谱技术进行水果成分特征分析,建立水果成分检测的属性集,然后根据属性的关联性进行信息熵赋予,采用决策属性辨识方法进行水果成分检测的规则向量集约简,最后采聚类处理方法进行水果成分的特征分类,实现水果成分检测。仿真结果表明,本文方法能够有效提取水果成分检测的特征,提高了水果成分检测的正确率。In order to improve the recognition and analysis ability of fruit components,a mathematical model of fruit composition detection based on near infrared spectroscopy is proposed in order to solve the problem of low accuracy of fruit composition detection.Firstly,the characteristics of fruit components were analyzed by near infrared spectroscopy,and the attribute set of fruit composition was set up.Then the information entropy was given according to the correlation of the attributes.The decision attribute identification method was used to carry out the rule vector intensive reduction of fruit component detection. Finally,the characteristics of fruit components were classified by clustering method.Class,fruit component detection.The simulation results show that this method can effectively extract the characteristics of fruit components detection and improve the accuracy of fruit composition detection.

关 键 词：近红外光谱技术 水果成分检测 数学模型 信息熵 

分 类 号：TN27[电子电信—物理电子学]