Inverse modeling and optimization of low-voltage power VDMOSFET´s technology by neural networks

In this paper an efficient optimization TCAD system is presented. Two neural networks, as TCAD system tools were created. The direct neural network, used for quick simulations in domain of observation, and inverse neural network, used for obtaining optimum values of input parameters necessary for manufacturing the device with desired characteristics. Using this TCAD system, successful optimization of low-voltage power VDMOS transistor was completed.