自适应杂交NGA及其在多值电路分析中的应用

NGA with Adaptive Crossover and Its Application in Nonlinear Circuit Analysis

摘要: 提出了一种具有自适应交叉算子的ＮＧＡ算法（简记为ＡＣＮＧＡ），以解决具有多模态函数的参数优化问题，ＡＣＮＧＡ能随着解空间参数结构的变化，自适应地选择不同的杂交算子并确定相应的杂交率，使解群在保持多样性的前提下，以较收敛率得到最优解集。该算法可求解非线性方程组，并主解具有任意约束关系的非线性电阻电路的所有解。

Abstract: A Niched Genetic Algorithm with adaptive crossover(ACNGA) is proposed.The modified algorithm can be used to solve the parameter optimization problems with multimode functions.With varying of parameter structure of solution space siutable crossover operators are adaptivly chosen and corresponding operation rates are detemined in order to maintain the diversity ofsolutions and get higher rate of convergence.ACNGA can solve nonlinear equations and nonlinear resistive circuits with arbitrary constitutive relation.