基于数据驱动的生产质量决策支持系统模型的研究

Research on data-driven based model of decision supporting system of manufacturing quality

基于生产过程实时监控系统获取的大量质量监测数据 ,提出了以数据仓库为核心 ,数据挖掘为驱动的企业生产质量决策支持系统模式。以粗糙集理论为主体技术进行数据挖掘 ,针对实际生产过程的特性 ,将经典粗糙集不可分辨关系进行扩展 ,引入相近和相似关系 ,可以有效处理实际生产过程中数据的背景噪音和数据不完备的状况。采用遗传算法解决粗糙集理论中知识简约的NP完全问题 ,挖掘出的各个生产过程的相互关联规则 ,通过人机一体优化策略 。

On the basis of large amount of qualitative monitoring data acquired from real-time monitoring system of productive process, a mode is proposed to enterprise's productive quality decision supporting system which takes data bank as a core and data excavation as a drive. Taking the rough set theory as a main technique to carry out data excavation and aiming at the characteristics of actual productive procedure to extend the nondistinguishable relation of classical rough set and introduce relations of similarity and resemblance, thus the situations of data background noise and data incompleteness in actual productive course can be dealt with effectively. The complete NP problem of knowledge reduction in the theory of rough set is solved by adopting genetic algorithm. By means of human-machine incorporate optimization strategy, the interrelated rules of each productive process being excavated will make policymakers gain the really needed knowledge.