基于自适应模糊神经网络的国毛毛条性能预测  

Prediction of Domestic Wool Top Properties Based on ANFIS

作　　者：李伟[1] 黄秀宝[1] 

机构地区：[1]东华大学纺织学院,上海200051

出　　处：《东华大学学报（自然科学版）》2006年第6期66-71,共6页Journal of Donghua University(Natural Science)

摘　　要：在对国产细羊毛原毛特征和毛条加工性能的研究的基础上,采用了基于自适应模糊神经网络(ANFIS)建模方法,分别建立了8个国毛毛条质量品质指标的预测模型。模型测试表明模型对毛条纤维的长度、长度变异系数、短纤维含量和加工过程精梳落毛率等指标均能进行较为准确的预测。With special reference to the characteristics and top-processing performance of domestic raw wool, a Neuron-Fuzzy Modeling method based on Adaptive -Network-Based Fuzzy Interference System （ANFIS） was established for predicting the indexes of wool top properties in this paper. The results show that the top fibre length, coefficient of variation of fibre length, short fibre content and combing noil can be predicted successfully.

关 键 词：国产原毛 加工性能 预测 ANFIS建模 

分 类 号：TS134.32[轻工技术与工程—纺织材料与纺织品设计]