摘要
根据坦洋工夫红茶感官品质主要评定指标,采用BP神经网络建立了坦洋工夫红茶等级与其主要评定指标之间的非线性映射关系,实现了对坦洋工夫红茶等级的预测,并通过实验进行验证。研究表明,所构建的坦洋工夫红茶评审专家预测系统有较好的识别效果,实用性强,克服了许多人为误差,提高了茶叶评审的速度,评茶人员只需具备一定的茶叶审评知识,就能正确地对坦洋工夫红茶的等级进行评定。
According to themain indexes ofsensoryqualityevaluation in the TanYangGongFublacktea, the grade of tea and nonlinear mapping relationshipin the main indexwasbuiltby usingBP Neural Network.Predictionof grade was achieved and all these have been proved by experiments.The results showedthat expert prediction system of TanYangGongFublack teapossessedgood recognition effectand strong practicality,couldovercomesubjective error caused by humanand to improvedthe speed of teaassessment,the tasters whopossessed theknowledgeof tea assessmentcouldcorrectly assessedthe gradeoftea.
出处
《茶叶科学》
CAS
CSCD
北大核心
2015年第5期465-472,共8页
Journal of Tea Science
基金
福建宁德市科技计划项目(20100011)
关键词
人工神经网络
坦洋工夫红茶
品质评定
预测系统
artificial neural network,Tan Yang Gong Fu black tea,quality evaluation,prediction system
