Research And Application Of A Modeling Method Based On Nonlinear Scattering Function And Wavelet Neural Network

Analysis and study of non-linear theory of Microwave circuit have gradually attracted people's attention. In this paper , based on the traditional non-linear characterization methods, a new expression of nonlinear scattering function is presented. Nonlinear Scattering function takes into account the generation of non-linear harmonic, thus can be used to analyze the non-linear characteristics of the circuit. Its definition, the physical meaning, nature and expression of networks connection are Focused on .After that, the method of black-box to model microwave devices in a large signal is proposed. there is no need to know the interior structure and work mechanism of the device, just based on the work characteristic of it's ports, we can emulate a equivalent model of parameters of the device. Artificial neural network is a kind of black-box model with rapid development recently. And the wavelet neural network which integrates the wavelet analysis and neural network closely has become an effective tool to solve non-linear modeling problem.The black-box model is based on frequency domain measurement. In this paper, we introduce how to measure devices in a real large signal situation, how to set up a test system of nonlinear scattering. Then a large amounts of data of calibration and test are obtained. Based on the test data of devices, nonlinear models are successfully established using wavelet neural network, then they are used to design amplifier.