Mathematical Modelling of the Bipolar Transistor using Artificial Neural Networks

Requirements for circuit simulation are increasing according to the development of system integration with many different functions. To assist the development, the most important modeling issue is to guarantee sufficient simulation accuracy and applicability for any advanced technology. For achieving this task it is inevitable to maintain a physically correct modeling of the real technology processes which govern the function of these BJT’s, even in the circuit simulation model. Here the approaches to realize the outlined requirements are summarized.