融合粗糙集和灰色系统的烟叶感官质量预测  

Sensory quality prediction of tobacco based on rough sets and gray system

作　　者：张光辉[1] 李俊丽[1] 陈海棠[1] Zhang Guanghui Li Junli Chen Haitang(College of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650504, Chin)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650504

出　　处：《计算机与应用化学》2017年第2期163-166,共4页Computers and Applied Chemistry

摘　　要：对于多因素影响的烟叶感官质量预测问题,本文构造了融合粗糙集和灰色系统理论的预测模型。该模型运用粗糙集的知识依赖度理论对多属性进行约简,在约简基础上建立感官质量多变量灰色预测GM(1,N)模型。用该预测模型对云南省某烟厂烟草产品的香气质质量进行了拟合和预测。实例验证结果表明:该模型具有较高的拟合、预测精度,为烟草感官质量预测问题提供了一种定量化方法,是对传统方法的补充和完善。A For multi-factor prediction problems of tobacco sensory quality, this paper constructs a prediction model integrating rough set with grey system theory. The degree of knowledge dependency is used to reduce attributes, and the multivariable gray forecasting GM （1, N） model is established based on knowledge reduction. Aroma quality of Yunnan tobacco is fitted and forecasted by using the established model. Verification results of example show that： the model has higher prediction accuracy and provides a quantitative method for predicting the sensory quality of the tobacco. It is a supplement and improvement to the traditional methods.

关 键 词：粗糙集 依赖度 属性约简 灰色预测模型 香气质 

分 类 号：TS411[农业科学—烟草工业]