Prediction of OLED Temperature Distribution Based on Neural Network

In recent years, the market share of OLED screens has increased year by year. Compared with traditional LCD screens, the advantages of OLED screens with bright colors and lower power consumption are attributed to the fact that LCD needs a backlight LED as the light source, and the pixels of the OLED screen emit light by themselves, so the power consumption of OLED is reduced and the thickness becomes thinner. In addition to the above reasons, OLED has a special advantage, that is, OLED can be bent, so the use of OLED screens is becoming more and more popular, and has even exceeded LCD screens usage rate. Although OLED has many advantages, it still has some disadvantages that need to be improved. OLED has a short lifetime. After a period of use, the luminance of the OLED will degrade, and the luminance degradation experienced by the different region of OLED screen is different, called differential aging, so it will cause burn-in. There have been some studies on OLED luminance degradation, including research on OLED luminance degradation under different temperature conditions, but they have not discussed the effects of temperature in differentareas of a panel. Therefore, this paper will discuss the temperature prediction for different regions of an OLED panel. The temperature prediction of the OLED panel is based on the four temperature sensors installed at the rear of the OLED panel and the picture displayed on the panel at the time. The prediction method is neural network that uses pre-collected data for model training. After the training is completed, the temperature distribution of the OLED panel is predicted based on the four temperature sensors and displayed pictures. The prediction error method uses mean square error, and the error value obtained is less than 0.2°C.