基于粗糙集的精毛纺粗纱工艺规则推理和案例组织  

Worsted roving process′s rule extraction and cases based on rough set

作　　者：刘贵[1] 于伟东[2,3] 

机构地区：[1]东华大学纺织材料与技术实验室,上海201620 [2]嘉兴学院服装与艺术设计学院,浙江嘉兴314001 [3]东华大学纺织学院,上海201620

出　　处：《纺织学报》2008年第12期17-21,共5页Journal of Textile Research

摘　　要：简要分析精毛纺前纺工艺的特点,针对其粗纱质量指标的重要性和涉及到13个工艺参数给企业工艺设计和生产质量控制带来困难这一问题,基于采集到的某精毛纺企业50组历史生产数据研究,对其实现粗糙集数据分析,在Rosetta环境下分别得到粗纱CV值(R1)和单重(R2)的约简核,分别提取出工艺参数和R1、R2之间的36条和28条潜在知识规则,并以其中某一批毛条为例,以约简后的属性组织案例与R1相关的第12条规则吻合;同时将粗糙集引入基于案例推理中,建立了基于粗糙集的案例库索引,实现了基于规则推理和基于案例推理的有效结合,为工艺设计和产品质量预报及控制提供指导。The character of worsted fore-spinning process has been briefly analyzed. As the roving quality＇s importance and involving many parameters, the quality＇s control and process parameters＇ design become difficult. Based on 50 groups of historical data gathering from a worsted mill, rough set data analysis （RSDA） was actualized to get the reduced cores of roving unevenness （ R1 ） and weight （ R2 ） in Rosetta environment. According to R1 and R2, 36 and 28 potential knowledge rules were got respectively. Take a certain group of tops as an example, the twelfth rule related R1 was completely in accord with the case composed by the reduced attributes. Meanwhile RS was inducted to case-based reasoning （CBR） to establish the case index storehouse. It realized the effective union of rule-based reasoning （RBR） and CBR, in order to guidethe craft design and product quality control.

关 键 词：精毛纺 前纺工序 粗纱 粗糙集 案例推理 

分 类 号：TS131.9[轻工技术与工程—纺织材料与纺织品设计]