基于极大熵差分进化混合算法求解非线性方程组  

Solving nonlinear equations based on maximum-entropy differential evolution hybrid algorithm

作　　者：陈海霞[1] 杨铁贵[1] 

机构地区：[1]郑州职业技术学院基础教育处,郑州450121

出　　处：《计算机应用研究》2010年第6期2028-2030,共3页Application Research of Computers

基　　金：国家自然科学基金资助项目(60875043)

摘　　要：针对非线性方程组,给出了一种新的算法——极大熵差分进化混合算法。首先把非线性方程组转换为一个不可微优化问题;然后用一个称之为凝聚函数的光滑函数直接代替不可微的极大值函数,从而可把非线性方程组的求解转换为无约束优化问题,利用差分进化算法对其进行求解。计算结果表明,该算法在求解的准确性和有效性均优于其他算法。To solve nonlinear equations, this paper presented a new method, referred to as maximum-entropy differential evolution hybrid algorithm. First, transformed nonlinear equations problems into non-differentiable optimization, and used the smooth aggregate function to instead of maximum function in non-differentiable optimization problem. Thereupon, the primal problem could be transformed into unconstrained optimization, then used the differential evolution algorithm to solve this problem. Numerical results show that the proposed method is more accurate and effective.

关 键 词：差分进化算法 非线性方程组 极大熵方法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]