近红外光谱污水COD分析系统设计  

Design of Near Infrared Spectroscopy Wastewater COD Analysis System

作　　者：李阳[1] 侯长宁 范日高 LI Yang;HOU Changning;FAN Rigao(Concord University College,Fujian Normal University,Fuzhou,China,350116;College of Electrical Engineering and Automation,Fuzhou University,Fuzhou,China,350108)

机构地区：[1]福建师范大学协和学院,福州350116 [2]福州大学电气工程与自动化学院,福州350108

出　　处：《福建电脑》2024年第6期82-86,共5页Journal of Fujian Computer

基　　金：福建省中青年教师教育科研项目(No.JAT220516);福建省自然科学基金项目(No.2021J01636)资助。

摘　　要：为了解决利用近红外光谱技术测定污水中的化学需氧量技术中样本数据集参数存在数据过拟合、干扰因素大、检测界面不直观等问题,本文采用一维卷积神经网络技术,建立水样近红外特征与污染指标成分含量之间的关系模型,并利用MATLAB的图形用户界面设计分析界面。实验结果表明,该方法预测结果良好,可实现水质污染的定性及定量的快速检测。In order to solve the problems of overfitting of sample dataset parameters,large interference factors,and non intuitive detection interface in the determination of chemical oxygen demand in wastewater using near-infrared spectroscopy technology,this paper adopts one-dimensional convolutional neural network technology to establish a relationship model between near-infrared characteristics of water samples and pollution indicator component content,and designs an analysis interface using MATLAB graphical user interface.The experimental results show that the method has good prediction results and can achieve rapid qualitative and quantitative detection of water pollution.

关 键 词：污水 近红外光谱 化学需氧量预测 

分 类 号：O657.33[理学—分析化学]