当前位置: 首页 > 文章 > 基于近红外透射光谱及神经网络的大豆油质量分析 湖北农业科学 2015,54 (1) 175-177
Position: Home > Articles > Quality Analysis of Soybean Oil based on Near Infrared Transmission Spectra and Artificial Neural Network Model Hubei Agricultural Sciences 2015,54 (1) 175-177

基于近红外透射光谱及神经网络的大豆油质量分析

作  者:
蔡立晶;蔡立娟;李文勇;赵肖宇;尚廷义
单  位:
大庆油田有限责任公司第一采油厂;长春理工大学;黑龙江八一农垦大学
关键词:
近红外透射光谱;BP神经网络;豆油质量分析
摘  要:
提出了一种基于近红外透射光谱及最速下降BP算法识别大豆油质量的方法。光谱采集范围是10 000~4 000 cm-1,将得到的近红外光谱数据作为网络的输入神经元,利用主成分分析方法得出8个变量指标数,该变量指标对样品累计贡献率达到99.9%以上;将8个主成分的特征值作为BP网络的输入向量,建立BP神经网络模型。该模型对预测样品集能正确判别,判别正确率达到100%。
译  名:
Quality Analysis of Soybean Oil based on Near Infrared Transmission Spectra and Artificial Neural Network Model
作  者:
CAI Li-jing;CAI Li-juan;LI Wen-yong;ZHAO Xiao-yu;SHANG Ting-yi;Heilongjiang Bayi Agricultural University;Changchun University of Sciences and Technology;The Fist Oil Production Company,Daqing oilfield;
关键词:
near infrared transmission spectroscopy;;BP neural network;;soybean oil quality analysis
摘  要:
A method based on near infrared transmission spectra and gradient descent BP algorithm was used to analyzed the quality of soybean oil. The range of 10 000 to 4 000 cm-1spectral was acquired, then the near infrared spectrum data was input to BP network. Eight variable indexes were obtained with principal component analysis. The cumulative contribution rate of the 8 variable indexes was more than 99.9%. Using the 8 index as input vectors of BP neural networks model, it can discriminate the quality of samples with the accuracy of 100%.

相似文章

计量
文章访问数: 8
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊