摘要
本文在遗传算法 (GeneticAlgorithm ,简记GA)与思维进化计算 (MindEvolution aryComputation ,简记MEC)的基础上 ,提出了一种广义进化模型 (GeneralizedEvolutionaryModel ,简记GEM)。该模型用微演化与宏演化两个过程 ,分别模仿人类的思维学习方式与自然进化 ,并通过概率趋同、信息迁移、自适应变异算子将两个过程有机的结合起来 ,从完全意义上模仿了人类的进化。该模型既能有效地克服遗传算法的本质缺陷 ,又能拓展思维进化计算的理论基础及应用范围。
Based on the Genetic Algorithm and Mind Evolutionary Computation, a new Generalized Evolutionary Model (GEM) consisting of microevolution and macroevolution is presented in this paper. Microevolution is introduced to imitate the thinking mode of mankind and macroevolution is introduced to simulate man's evolution in nature. At the same time, some operators are designed to integrate two kinds of evolution, such as probability similartaxis, information migration , self-adapting mutation, ect. GEM not only can overcome GA's intrinsical flaws effectively, but also can broaden MEC's theory and the area of applications. Numerical optimal simulations show that GEM is effective.
出处
《太原重型机械学院学报》
2002年第3期181-186,共6页
Journal of Taiyuan Heavy Machinery Institute
基金
863智能计算机系统主题与国家自然科学基金资助 (836 - 30 6 - 0 6 - 0 6 - 66 0 174 0 0 2 )
关键词
广义进化模型
遗传算法
思维进化计算
趋同
异化
genetic algorithm
mind evolutionary computation
similartaxis
dissimilation