Genetic algorithm techniques for the design of nonlinear microwave circuits

Abstract

Steady state analysis of nonlinear circuits represents the most computational challenging problem in RFMcrowave circuit design. Even though all circuits are assumes as a linear circuits in circuit theory, in fact it is nonlinear circuit. The problem in designing and analyzing the nonlinear circuits usually more complicated than for linear circuits. Many approaches have been done to overcome this problem. Recently hybrid method of analyzing nonlinear circuits has proved to be popular. In this report a hybrid method of Sample Balance Genetic Algorithm (SBGA) used to analyze nonlinear diode excited by RF circuit. The purpose of this report is to minimizing the current difference between nonlinear current which across the nonlinear element and linear current at linear element and voltage across the nonlinear terminal. By using Sample Balance's differential of current linear and nonlinear equation as an objective function, a Genetic Algorithm routine then been constructed. Genetic Algorithm will then recombine, mutate and evaluate the population of solution to find the best solution based on Sample balance equation. This hybrid method manages to get 0.013A of error current and a sinusoidal waveform clipped at 0.W during first half of the wavefonn for the nonlinear voltage.