CAJ | 학술논문

目前在普洱茶生产过程中的质量控制一环，以质量检验人员的经验为基础的感官品评仍然占据主导地位，其主观性强、准确性和重现性较差的缺点逐渐暴露出来。针对这一难题，傅里叶变换近红外光谱技术结合适当的谱图预处理方法和聚类分析算法，实现了对多种普洱茶样品聚类与定性分析，模型判别准确率能达到90%以上，稳定性与重现性极高。该方法被证明在普洱茶的品质分析与质量控制方面具有较大的应用价值和乐观的应用前景。
목전재보이다생산과정중적질량공제일배，이질량검험인원적경험위기출적감관품평잉연점거주도지위，기주관성강、준학성화중현성교차적결점축점폭로출래。침대저일난제，부리협변환근홍외광보기술결합괄당적보도예처리방법화취류분석산법，실현료대다충보이다양품취류여정성분석，모형판별준학솔능체도90%이상，은정성여중현성겁고。해방법피증명재보이다적품질분석여질량공제방면구유교대적응용개치화악관적응용전경。
@@@@A series of problems have occurred in quality control during the manufacturing process of Pu-erh since sensory evaluation, which mainly based on the experiences of QCs, is still the kernel technology. Its serious subjectivity, poor accuracy and bad repeatability are gradually exposing. FT-NIR, combined with appropriate preprocessing methods and classification analysis algorithms , has achieved the classification and qualitative analysis of a series of Pu-erh samples , identification accuracy rate could reach 90%or higher. Thus the method is proved to be of great value and optimistic application prospects in the quality analysis and control of Pu-erh for its good stability and repeatability.