一种基于遗传算法的求代数方程组数值解的新方法  

New method based on genetic algorithm for resolving algebraic equation groups

作　　者：胡小兵[1] 吴树范[1] 江驹[1] 

机构地区：[1]南京航空航天大学自动控制系,南京210016

出　　处：《控制理论与应用》2002年第4期567-570,共4页Control Theory & Applications

摘　　要：解方程组是工程研究中的基本问题 .当前的各种求解方法都只针对某一特定类型的方程组 ,没有通用性 .根据遗传算法与具体问题无关的特点 ,提出了一种通用的代数方程组数值解求解方法 .首先 ,基于遗传算法建立了代数方程组求解问题的通用模型 ;然后阐述了具体的求解过程 ;最后分析了该法的特点和性能 .实验表明该法是有效可行的 ,进一步提高解精度的关键在于对遗传算法本身的研究 .Resolving equation group is a principal problem in engineering study. Any existing method can only deal with a specified kind of equation group, so has no broad applicability. A new method with broad applicability is presented on base of genetic algorithm, because genetic algorithm does not at all care what the problem is. First, a common model is established for all kinds of algebraic equation groups; then a software process is brought up; finally the performance is analyzed. The results of experiment show that the new method is effective and feasible, and the key to improve this method is to study genetic algorithm further.

关 键 词：遗传算法 代数方程组 数值解 适应度函数 聚类 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]