论文部分内容阅读
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。
To predict the sensory quality of tobacco leaves with multiple factors, this paper constructs a prediction model based on the fusion of rough sets and gray system theory. The model uses the theory of knowledge dependence of rough sets to reduce the multiple attributes, and establishes a multivariate sensory gray prediction GM (1, N) model based on the reduction. The predictive model was used to fit and predict the aroma quality of tobacco products in a cigarette factory in Yunnan Province. The experimental results show that the model has a high fitting and prediction accuracy, and provides a quantitative method for the prediction of tobacco sensory quality, which is a supplement and improvement to the traditional method.