摘要
该文设计了一个面向质量管理工程专业学生的综合快检实验。实验以便携式光谱快速采集样品质量特性数据,运用质量统计工具和智能学习算法实现数据快速分析和规律揭示。通过拉曼光谱仪照射乳品获得样品的散射特征信号,每个实验样品数据采集仅需60s。选取特征峰值,结合质量统计控制分析工具绘制样品质量控制图,可描述样品质量波动情况。运用极限学习机算法结合数据预处理方法,优化实现样品品牌快速判别,识别率可达97.3%,算法运行时间仅需1s。
A comprehensive rapid inspection experiment for students majoring in quality management engineering is designed.The experiment uses a portable spectrometer to quickly collect sample quality characteristic data,and then uses a quality statistical tool and an intelligent learning algorithm to achieve rapid data analysis and pattern revelation.By irradiating dairy products with a Raman spectrometer,scattering characteristic signals of the samples are obtained,and the data collection for each experimental sample only takes 60 seconds.Feature peaks are selected and quality statistical control analysis tools are uesd to draw a sample quality control chart,which can describe the fluctuation of sample quality.Furthermore,by combining an extreme learning machine algorithm with data preprocessing methods,the rapid identification of sample brands can be optimized,with a recognition rate of 97.3%and an algorithm running time of only 1 s.
作者
张正勇
ZHANG Zhengyong(School of Management Science and Engineering,Nanjing University of Finance and Economics,Nanjing 210023,China)
出处
《实验科学与技术》
2025年第1期12-16,共5页
Experiment Science and Technology
基金
南京财经大学教学改革项目(JGY202270,JGZ2023003,JGY2023081)
南京财经大学高等教育及改革发展研究项目(GJGF202136)
南京财经大学《仪器分析概论》产教融合一流课程建设项目(南财教字[2023]88号)
江苏省高等教育学会高校实验室研究委员会项目(GS2022BZZ19)
江苏高校“青蓝工程”资助项目(苏教师函[2021]11号)。
关键词
实验教学
质量快检
质量控制
质量分析
智能识别
experimental teaching
quick quality inspection
quality control
quality analysis
intelligent recognition
