Research On Real-time On-line Intellegence Quality Control Technology For The Tobacco MES

MES,an execution system lying betwwen management level and workshop control,can achieve the effective integration of enterprise production activities and business activities,optimize the production operation,the quality control and the enterprise management,and thus becomes one of the new technologies worthy of tobacco companies'close attention.Meanwhile,as the eternal theme for the enterprise, quality is the magic weapon for the enteprise to realize the sustainable development and to win the competition in the future.Real-time online quality control is the main resaech area and developing trends of the modern control technology.Based on the MES platform, the tobacco quality control technology is the main reason for the rapid developlment of the tobacco companies,which make it to be the research fronts and important topic in the studies of the tobacco company.In this thesis, the author studies the theory of the real-time online quality control technology of the tobacco company and its implementation methods. This thesis mainly including the following several parts:(1)The author fist studies the basic theory of the MES and intelligent quality control.Based on this, the author then analyzes the meaning of the real-time online quality control system, recognizes the basic flow of it and finally sets up a quality control pattern and system structure based on the neural network.(2)To meet the need of the real-time monitoring of the parameters of the quality charateristics, the author improves the calculating methods of simulated annealing algoritm and Elman neural network, and establishes the intelligent quality control and recognize system based on the improvement of Elman network using simulated annealing algoritm , which features with rapid learning and rapid recognition, and can be applied to the real-time online dynamic processing and the monitoring of the quality charateristic parameters. (3)After the study of the recognition of the abnormal pattern of the control chart of Elman network,the author has put forward can recognizes the basic pattern of the control chart and the 5 kinds of abnormal pattern, which provide the technological support to theintelligent quality control system based on SPC. In this thesis, the author adopts the PCA technologies to deal, which shortens the pattern identifying time and improved the identification accuracy. And this can be applied to the real-time online pattern indentification of the control chart.(4)The ahuthor analyzes the abnormal cause diagnosis in the real-time online intelligent quanlity control technologies and puts forward the cause and effect diagnostic techniques based on the improved neural network. And based on the abnormal pattern recognition, the result of recognition will activate the cause diagnosis system directly. And this shortens the calculating time of the system, improves the realtime of the system. And this can be applied to the real-time online cause diagnosis of the quality control.The effect way of the intelligent quality control abnormal pattern recognition and cause diagnosis this thesis has put forward adapts to both simple and complex production quality control system. What this thesis has studied has important theoretical and applied value to the real-time, automated and intelligent quality control of modern manufacturing system.