Study on the Continuous Quality Improvement Systems of LED Packaging Based on Data Mining

LED is one of the most widely used components in electric products. And the LED packaging is a very important process between semiconductor manufacturers and the electric product manufacturers. Based on the analysis of the characteristics of the LED packaging processes, a quality control model based on SPC (statistical process control) and data mining is put forward. The data mining is used as the quality data analysis tool and the quality diagnosis method. Then an infrastructure of the integrated continuous quality improvement systems of the LED packaging is put forward. In this infrastructure, there are three layers of the data collection layer, the data analysis layer and the result viewer layer. Furthermore, the data warehouse of LED packaging is designed with the snowflake schema. A 3 layer yield rate SPC is studied and the decision tree method is used as a quality diagnosis method based on the designed data warehouse. Finally, a prototype of the continuous quality improvement system is developed.