Utilization of Process Variation for System Evolution

The basic method of statistic process control is to eliminate variation and put process under statistic control. The negative effect of the method is that it leads to the neglect of the valuable information contained in process variation, thus missing the opportunity for process optimization. But variation is not necessarily harmful, from the biological evolution point of view, variation is the driving force of evolution. Inspired by biological evolution mechanism, a method of adaptive optimization based on process variation is proposed. A data mining technology - association rules mining technology is adopted to analyze the SPC data and excavate association rules in the quality parameters. Then the association rules are evaluated and utilized to improve production processes. The method expand the stability-oriented quality control system to stability-optimization- combined system and continual quality management improvement is realized by iteration of this data driven process. A case study of injection molding process optimization is provided to illustrate application of the presented method.