摘要
提出了一种新型解空间种群均匀化的自适应遗传算法,并采用随机方法对初始种群加以改进,使初始种群均匀分布于解空间之中.在优化进程中,引入自适应算法,使交叉和变异算子具有自适应性;将自适应调节机制引入适应值函数中,使适应值函数同样具有自适应性.为证实所提出的改进遗传算法的可行性和有效性,对几种典型的多峰值函数进行了寻优测试.优化测试结果与解析解及标准遗传算法优化结果相对比,证明改进遗传算法的全局搜索能力和收敛性都远优于标准遗传算法.
An improved self-adaptive genetic algorithm （GA） with population uniformity in the whole solution space has been proposed, and the initial population has been improved using the random method for increasing convergency of the optimum in the global space. In optimizing, self-adaptive algorithm has been introduced to adjust the self-adaptive cross operator, self-adaptive mutant operator and fitness function. For verifying the feasibility and validity of the proposed GA, some typical nonlinear testing functions with multiple extremes have been testified. By comparison of the optimized testing results with the analytical solutions and the traditional ones, it is proved that the global searching ability and convergency speed of the proposed GA are higher than those of the standard genetic algorithm （SGA）.
出处
《沈阳工业大学学报》
EI
CAS
2005年第6期623-628,共6页
Journal of Shenyang University of Technology
基金
辽宁省博士科研启动资助项目(20041026)
关键词
遗传算法
自适应
初始种群
遗传算子
优化
genetic algorithms
self-adaptation
initial population
genetic operators
optimization
