使用遗传算法进行微波非线性电路包络仿真

Applying Genetic Algorithm in Nonlinear Circuit Envelope Simulation

介绍用于分析任意调制信号激励的微波非线性电路包络仿真技术,相应给出了线性器件的伴随模型,并将该方法发展为分析多载波调制信号激励的强非线性电路。此外,本文给出了一种求解非线性谐波平衡方程组或交调波平衡方程组的非数值方法———基于十进制编码的混沌遗传算法(DCGA),融合设计了相应的交叉和变异操作,并将混沌过程引入其中以避免陷入早熟状态,可有效地产生更优解,并加速种群的收敛。最后,给出了运用这种方法的两个计算实例。

A new kind of microwave nonlinear circuit analysis method, the circuit envelope （CE） simulation technique under arbitrary modulated signals excitation, is introduced in the paper, The corresponding concomitant models of the linear devices applied in the CE method arc also discussed. Then, the method is developed to obtain the responses of the strong nonlinear circuits excited by multi-carrier modulated signals. On the other hand, a novel decimal-coding chaos genetic algorithm （DCGA） combining with the suitable crossover and mutation operations is firstly presented for solving the above nonlinear harmonic balanced equations or inter-modulation balanced equations. The chaos process is used to avoid premature states in the whole evolution progress. Finally, two examples are studied to testify the validity of the method in the paper.