Network Model for Parametric Synthesis of Resonant-Tunnel Heterostructure

The resonant-tunnel diode is a promising element for application in various radio-electronic systems, which is due to the possibility of searching for the optimal combination of electrical characteristics of the resonant-tunnel structure and its design and technological parameters at the design stage. The results of solving the problem of parametric synthesis of the resonant-tunnel structure are presented. To reduce the iteration process of synthesis it is necessary to use a fast-acting model of current transfer, for this purpose a model based on an artificial neural network is used. To train the neural network model, the results of numerical simulation were used. In order to improve the accuracy of the neural network model, an ensemble method (bagging) of machine learning was used. The use of the model based on neural networks allowed us to achieve high accuracy and speed required for parametric synthesis of the resonant-tunnel structure.