Design And Implementation Of Chatter Marks Monitoring System For Continuous Annealing Line Based On FA-KICA

With the rapid development of computer and electronic measurement technology, it brings higher integration and higher level of modern industrial process automation. It also brings higher requirements to the industrial production process monitoring. Process monitoring of industrial production was mainly on the production process or the field instrumentation. By studying the mechanism of the production process, combined with the principle of mathematical methods and control science, it established an effective monitoring model to monitor the production and production of a certain degree of warning. This can improve the efficiency and quality of products.Fault diagnosis method based on data does not need to create complex mathematical models and the accurate prior knowledge, the data is first-hand information on industrial processes, and it is closer to the true situation, suitable for complex industry. PCA is one of the most typical methods. Its principle is to find a new variable to replace the original variables, which carried on the maximum information of the original data, by reducing the dimension of the original data. It will be easier for diagnosis.The chatter marks of cold rolled flat strip seriously affect the vibration pattern of the product quality, this paper is based on the background of chatter marks, through the depth analysis of the relationship between the mechanism and data, to identify the closely variables related with the formation of defects in the process. This paper is based on the PCA method, discussed the KPCA and MFA-ICA methods, compared their strengths and weaknesses, and then choose the most suitable method based on the actual formation of the monitoring mechanism. This paper also established relationship mode between chatter marks and the state of chatter mill, diagnosed process variables or parameters that lead to formation of chatter marks, and given the guidance proposed production operations. Finally, it designed system database and user interface to ensure that the fault monitoring system online.In this paper, the fault monitoring system for chatter marks is running very effective in the actual system, after evaluation, it can effectively diagnose on-site system failure and human error in the process. This ensures the normal operation of industrial formation and to provide the basis for later study.