基于Metropolis判别准则的遗传算法

Genetic Algorithm Based on the Metropolis Criteria

针对遗传算法在应用过程中出现的过早收敛问题，引入Ｍｅｔｒｏｐｏｌｉｓ判别准则对复制算子加以改进，并从理论上对遗传算法的收敛性进行分析。仿真结果证明了该算法解决过早收敛问题的有效性。

For solving the premature convergence problem, the paper presented a new reproduction operator by combining the optima-maintaining with the Metropolis criteria, and analysed the convergence property of the modified genetic algorithm. The simulation results demonstrated the validity of the modified genetic algorithm solving the premature convergence problem.