面向微波组件工艺失效分析的大数据建模技术

快速、准确定位微波组件生产过程中的质量问题，对提升微波组件的质量可靠性、工艺稳定性以及生产效率具有重要作用。在微波组件工艺质量问题传统人工分析逻辑基础上，通过对当前微波组件生产流程各环节的数据特点的挖掘分析，提出了生产大数据与失效分析知识融合的建模方法，并应用于工艺问题的辅助排故中。首先，基于微波组件工艺质量数据特征进行数据清洗得到故障关键数据，作为大数据挖掘建模的基础数据；其次，从微波组件质量特征相似性的角度对不同微波组件进行聚类处理，提升稀疏数据的信息密度；最后，采用大数据挖掘算法融合失效分析先验知识建立用于辅助排故的知识模型，并基于样本数据对提出的建模方法进行了实例分析和模型的软件化部署，验证了在微波组件工艺质量问题分析应用中的可行性。

Big Data Modeling Technology For Microwave Component Process Failure Analysis

It plays an important role in improving the quality reliability, process stability and production efficiency of microwave components that rapid and accurate positioning of the quality problems in the microwave components production process. Based on the traditional manual analysis logic of microwave components process quality problems, through the mining and analysis of the data characteristics of every link in the current microwave components production process, the modeling method of production big data and failure analysis knowledge fusion is put forward, and it is applied to the auxiliary troubleshooting of process problems. Firstly, based on the characteristics of microwave components process quality data, the key fault data is obtained through data cleaning, which serves as the basic data of big data mining modeling. Secondly, from the perspective of quality feature similarity of microwave components, cluster different microwave components to improve the information density of sparse data. Finally, the big data mining algorithm is used to fuse the prior knowledge of failure analysis to build the knowledge model for auxiliary troubleshooting. Based on the sample data, the case analysis and the software deployment of the model are carried out for the proposed modeling method, which verifies the feasibility of the application in the analysis of microwave component process quality problems.