大数据赋能A铝厂首个五层一体钎焊板工艺开发落地

Big Data Assistance to Development of Five-Layer Brazing Plate Process in A Aluminum Plant

首先对三层板轧制时设备、生产、工艺和质量数据进行实时采集存储,构建三层板轧制数据模型和工艺知识库;在此基础上,试验开发新五层板产品,利用三层板知识库确定监控工艺参数,持续收集工艺过程数据并进行清洗、编码与关联特征提取等预处理方法获得训练/测试数据集;采用神经网络建立新工艺质量在线预测模型,对工艺参数进行评估与优化;通过应用验证表明该方法加快了新工艺落地和持续改善。

In this paper,the equipment,production,process and quality data of three-layer plate rolling are collected and stored in real time,and the three-layer plate rolling data model and process knowledge base are constructed.On this basis,a new five-layer board product is developed,the three-layer board knowledge base is used to determine the monitoring process parameters,the process data is continuously collected,and the training/test data set is obtained by pre-processing methods such as cleaning,coding and correlation feature extraction.