摘要
分析了导致进化策略早熟收敛的原因,提出了一种新的双种群进化策略算法,进化分别在两个不同的种群间并行进行,两个种群采用不同的变异算子。将该算法用于求复函数方程的解,该方法具有计算精度高、自适应性强等特点,最后的实例表明该算法优于传统的迭代法、下山法等方法。其目的为求复函数方程的根给出一新算法,该算法在科学与工程计算中有着重要地应用。
Premature convergence is the fatal shortcoming of the traditional evolution strategy.Based on the analysis of premature convergence,a novel hi-group evolution strategy algorithm is proposed,In the new algorithm,evolution of the two groups is parallel performed with different mutation operator respectively.Using the new algorithm to find all roots of complex functional equation, this algorithm has much character such as the solution is in high accuracy,strong adaptability and so on.At last from the examples the authors can know that the novel algorithm is better than the old method such as iterative method,down mountain and so on.The object is to propose a new algorithm for solving complex functional equation.This algorithm has very important application in the computation of scienee and engineering.
出处
《计算机工程与应用》
CSCD
北大核心
2008年第7期78-81,共4页
Computer Engineering and Applications
基金
国家自然科学基金(the National Natural Science Foundation of China under Grant No.60461001)
广西省自然科学基金(the NaturalScience Foundation of Guangxi Province of China under Grant No.0542048)
广西研究生教育创新计划资助项目(No.2007106080701M18)
广西民族大学重大项目资助课题
关键词
双种群
进化策略
复函数方程
高斯变异
柯西变异
bi-group
evolution strategy
complex functional equation
Gaussian mutation
Cauchy mutation
