改进的遗传退火算法在针对含有超越函数的非线性方程组中求解的应用

Improved Genetic-annealing Algorithm's application in solving the nonlinear equations which Contains transcendental function

传统非线性方程组解法在求解含有超越函数的非线性方程组中常常会遇到初始值敏感、收敛性差等问题,而鱼群算法、耦合神经网络算法等进化算法则有求解精度低等缺陷.考虑到该类方程组的特性,本文提出一种基于遗传退火算法来求解的进化算法,数值仿真实验结果表明,此方法在求解过程中不仅克服了传统方法中存在的初始值敏感、收敛性差等问题,并能求出精度颇高的解(精度远远高于鱼群算法、耦合神经网络算法),从而为该类方程组提供了一种高效的进化求解的方法.

High sensitive to the initial value and bad convergence reliability,which are the most problems when taking classical algorithm to solve nonlinear equations which Contains transcendental function.And also There are some limitations such as poor precise for the answer in AFSA's algorithm and CNN's algorithm.To aim at such problems,a method base on Genetic-annealing Algorithm will be introduced in this paper. The computer simulated results shows that this method not only avoid the problems above,but also make the result more accurate(more accurate than Artificial-Swarm Algorithm,Neural-Network Algorithm),and it come to be a new method for resolving such equations.