Research on Nonlinear Response Modeling Methods using Volterra Series and Neural Network model

Nonlinear response modeling plays an important role in the performance analysis of electronic systems. Electronic systems are usually composed of a large number of electronic components. When establishing a circuit-level system response model, a large number of device parameters need to be considered, and the simulation is very time-consuming. In addition, hardware parameters are difficult to measure accurately, which limits the application of modeling methods based on device parameters and circuit structures. This paper proposes a nonlinear response modeling method based on low-complexity Volterra series modeling and recurrent neural network modeling. The basic principles of this method are given, a simulation system is built, and the performance comparison analysis of different methods is carried out, providing important technical means for nonlinear response modeling.