Parametrization Techniques Applied to Filters Design
暂无分享,去创建一个
This paper describes different numerical methods for filters design and parameterization. Each technique is used for establishing an accurate mathematical model that links the circuit electromagnetic (EM) performances to the structure studying parameters. The construction of such mapping between the circuit parameters and its response is started from a limited number of rigorous time or frequency full wave simulations. Once the parametric model built, it can be used for evaluating the microwave circuit response as its parameters are being changed without any need of solving the Maxwell equations or additional meshing. The presented techniques are applied in both time and frequency domains. The first method using artificial neural networks (ANN) is used for studying a CPW microwave filter. Then, a hybrid technique combining ANN and infinite impulse response (IIR) filter is used for modeling a microstrip filter in the time domain. Finely, a cavity filter is studied in the frequency domain using a pole expansion technique.
[1] Paolo Arcioni,et al. Combination of generalized admittance matrices in the form of pole expansions , 1999 .
[2] V. Fouad Hanna,et al. Neural networks modeling and parameterization applied to coplanar waveguide components , 2000 .
[3] K. Kurokawa,et al. An introduction to the theory of microwave circuits , 1969 .
[4] M.F. Wong,et al. Combining neural networks and IIR filters for circuit microwave design and optimization in the time domain , 1999, 1999 IEEE MTT-S International Microwave Symposium Digest (Cat. No.99CH36282).