Study on the ink flow model of Lithographic Printing

This paper presents a method to forecast ink flow characteristics of Lithographic Printing. As the movement of the vibrator rollers of the ink system, the output ink from one ink zone of the fountain roller will spread to multiple ink zones on paper. The ink distribution on the printing paper is not only influenced by image coverage of the current zone but also the nearby zone. This paper collects sample data of ink distribution on the paper which is produced by computer simulation method when one ink zone output a unit quantity ink to train BP neural network. After reading the actual image coverage of the digital files we can use the trained BP neural network to build a matrix which reflects ink flow characteristics. By this matrix model we can predict the distribution of ink on paper when a certain output ink is given. It is significant for accurately adjusting and controlling the output of ink amounts in each ink zone.