【摘要】 提出了一种新型的粗粒度并行遗传算法 ( CGGA) ,该算法利用多个子种群基于不同的编码方式进行进化计算 .首先各子群体独立进行交叉、变异和选择遗传操作 ,每代进化后迁移算子被引入用来进行种群间的信息交流 ,迁移算子将各个子种群的最优个体替换相邻种群最差个体后继续进化 .基于时齐遍历马尔可夫链理论 ,给出了 CGGA各个子种群的概率转移矩阵与其进化概率转移矩阵 ,证明了以概率 1全局收敛 .对典型的测试函数 CGGA进行了求解 .仿真结果表明 ,本算法的收敛性能优于经典遗传算法 ( CGA) ,可以有效解决 CGA的过早收敛问题更多还原

【Abstract】 Coarse grained parallel genetic algorithm(CGGA) applies multi population to optimize the objective problem. Different population is based on different coding schemes. The migration factor is introduced to the CGGA. The respective population’s elite individual replaces the neighboring population’s worst individual. The search process of CGGA is an ergodic homogeneous Markov chain. The global convergence proof of CGGA was presented. The simulation on the benchmark problems shows that the CCGA achieves better convergence performance than canonical genetic algorithm(CGA). The inherent premature convergence problem can be greatly levied by the CCGA.更多还原