The Application Of Multivariable Statistic Process Control For The Process Of Printing Plate Production

Multivariable statistic process control method can be applied for printing plate manufacturing processes due to multi- nonlinearity and correlation of the large number of the process variables which have common influence on coating quality. the paper gives an overview of the methods and its application based on multivariate statistical projection methods (Principal Component Analysis, PCA). and the model of principal component is being build up by extracting principal characters from the multivariate manufacturing process of the printing plate. It can cut down the number of the variables monitored in the system. Furthermore it can also be used for analyzing, monitoring and diagnosing the production processes if the system is under control or not. As to the abnormal process, the root causes can be defined by the contribution of the variables to the statistics.A multi-variable model was also set up that combined coating quality and sensitivity with the main effected variables in the coating process. This regression method can be used to predict the coating characters with the setting of the process variables for continuous improving of the production processes.