摘要
针对传统遗传算法收敛早熟问题,在传统包含模拟退火的混合遗传算法的基础上,设计加入了"包含浓度均衡措施的复制算法",通过调整轮盘赌扇区面积,防止个体适应度的两极分化。从而避免了算法过早收敛于局部最优解;同时通过一个工程实例计算验证了算法的可行性。
For the premature convergence in traditional Genetic Algorithm, this paper comes up with a new enhanced algorithm which introduced new strategy called ＂equalization measurement on the concentration of the replication＂, together working with the traditional hybrid SA-GA. Through dynamic adjusting on roulette sector to prevent polarization of individual fitness, so that avoiding premature convergence of the algorithm to local optimal solution; Mean while validated the feasibility of algorithm with one engineering example.
出处
《计算机系统应用》
2011年第10期224-227,共4页
Computer Systems & Applications
关键词
收敛早熟
模拟退火
混合遗传算法
premature convergence
simulated annealing
hybrid genetic algorithm
