基于一致性度量的在线产品品质稳定性评价

Quality Stability Evaluation for Online Product Based on Conformity Measure

烟叶近红外光谱承载了品质信息的综合体现,是快速进行品质检测与评价的有效技术手段。基于特征光谱采用一致性检验方法建立了品质稳定性评价的一致性度量模型,能够有效筛选品质异常的烟叶样品,模型测算的CI值与感官评吸专家打分结果具有很好的相符性,二者的相关系数为0.95。试验结果验证了该模型的有效性,全过程自动在线完成,该方法对烟叶复烤加工过程品质控制具有较好的指导作用和实用价值,对推动我国智能制造技术的发展有重要意义。

The near infrared spectra of the tobacco carries the comprehensive information of quality,which is an effective technique for the rapid detection and evaluation of the product quality.In this paper,the conformity model is established based on characteristic spectrum for the evaluation of the quality stability.The model can discriminate outlier samples effectively.The CI value predicted by this model is consistent to the sensory evaluation,and the correlation coefficient is 0.95.The experiment result showed the effectiveness of the model,and the whole process is accomplished automatically on line,which has good guidance and practical value for the quality control of tobacco leaf redrying process.The study is significant to the development of the intelligent manufacturing technique in China.