Design and implementation of an intelligent quality prediction system

The purpose of this research is to design and implement an intelligent prediction system for manufacturing quality. 12 manufacturing factors are chosen from four categories of MEPH (Material/Equipment/Product/Human), through the sampling, features extraction, or features conversion system to present features of manufacturing system in this research. Then, we take quality samples composing by 12 features and the 16 quality assessment factors obtained to process system modeling by using the intelligent method of back-propagation neural network. After system modeling, these un-estimated manufacturing quality factors, such as materials, machines, products and people, etc., can be evaluated their quality assessment immediately. However, the 16 quality assessment factors obtained can help us to achieve the total quality assessment through advanced data fusion, then to get the prediction result of quality. Finally, this paper will propose the design and implement of the software and hardware archtecture, and take a lamination plant as an example, to carry out the expermient of manufacturing system quality prediction. The good testing results will prove the feasibility of this research.