The predictions of optoelectronic attributes of LED by neural network

In this paper, the predictions of optoelectronic attributes of Light-Emitting Diode (LED) chip, including luminous intensity, wavelength and forward voltage by using neural network were presented. The simulated data was measured by Electrical Luminescence (EL) technique. The well-trained neural models were used to predict the optoelectronic attributes of LED chip in its epitaxy growth stage in advance. These predicted results could provide the necessary information for the process engineer to adjust the control parameters of epitaxy growth accurately and then ensure the LED chip to be in conformance with the requested quality.