摘要
模拟了几种常用饲料原料在卧式混合机中的混合过程,通过美国ASD公司的QualitySpec5070近红外光谱仪全程采集了样本的近红外光谱数据,对样本数据进行了适当的预处理与波段选择,并建立了3种不同的有监督定性分析模型来判别饲料混合均匀状态。结果表明,模型的预测准确率均大于85%,为快速、准确、稳定地检测饲料混合均匀度提供了新的途径。
The experiment studied using the near infrared spectrum technology for feed mixing uniformity determination , I have mixed several feed raw materials through the small horizontal mixer and gathered the samples of real -time spectral data , preprocessed the data and made the qualitative analysis of pattern recognition ,then compared the prediction effects of methods .In the end I summed up the three kinds of the data processing methods to provide a feasible and convenient method of feed uniformity testing for future .
出处
《农机化研究》
北大核心
2015年第3期191-194,共4页
Journal of Agricultural Mechanization Research
基金
国家自然科学基金项目(50865005)
