摘要
基于事例的推理(CBR)和人工神经网络(ANN)技术,建立了精毛纺纱线质量智能预报加工模型。通过CBR,根据所加工产品的主要特征,能够从历史数据库中快速提取出与设计特征参数最为相似的案例,相似性高的案例排在检索案例前面。利用ANN预报模型对所调整方案的效果进行评价,一方面可以通过调整相应的加工工艺参数及产品质量指标,预测所加工纱线的质量指标,并与实际要求比较,确定最终的加工参数;另一方面,根据所要加工纱线的质量要求,反演所需毛条的相关质量指标,以便企业能够用较低的原料成本生产同样质量的纱线。
In this paper, the intelligent prediction model for worsted yarn manufacturing is huilt and researched using CBR and ANN. Through CBR, the most similar cases with the main design characteristics of the product required can be rapidly retrieved from past and vast eases,which are sequenced according to the value of the similarity degree. Then ANN model is used to compare and assess the predicted results after adjusting the corresponding process parameters, and the process technics can be optimized. On the other hand, the properties of wool top are also deducted using the prediction model through the characteristics of yarn required, which benefits to reducing the cost for the mill.
出处
《毛纺科技》
CAS
北大核心
2006年第5期5-8,共4页
Wool Textile Journal
基金
国家技术创新项目(02CJ-14-05-01)
