Prediction Method of the Paper Media Temperature in Electrophotographic Process Printing by Using LSTM Network

Abstract

The objective of this study is to predict paper media temperature at paper outlet in electrophotographic printing in order to improve the toner fixing quality. Continuous printing experiments were conducted under various conditions of printing speed, toner quantity, paper type, etc. using an electrophotographic printing machine. The temperature of the printer components and the temperature of paper at paper outlet were measured. Using these data, a prediction model by LSTM, which is one of the neural networks, was trained. From this model, the temperature of paper at paper outlet was predicted from the printer components temperature data at several sheets in the initial stage of printing, and it was compared with the measured value.