改进自适应遗传算法研究及其应用  

Research into and Application of An improved Adaptive Genetic Algorithm

作　　者：闻骥骏[1] 沈典栋[1] 沈成武[1] 

机构地区：[1]武汉理工大学,武汉430063

出　　处：《交通与计算机》2003年第1期3-5,共3页Computer and Communications

基　　金：国家自然科学基金项目资助(19772038);湖北省交通厅科技项目资助(鄂交科教[2002]46号)

摘　　要：针对标准遗传算法收敛慢,提出了一种改进自适应遗传算法,即通过全局变异算子和局部变异算子共同作用。改善种群的分布特性,加强算法的全局收敛能力。应用概率模型的遗传算法对4个著名的测试函数进行优化计算,实验结果表明:该算法不易陷入局部极值,收敛速度快,且实现简单。Considering the deficiency of standard genetic algorithm, we has introduced an improved adaptive genetic algorithm, which can improve the population diversity by both the global mutation operator and local mutation operator. It has increased the algorithm's capacity of global convergence. Four famous test functions of the optimization method are calculated with PGA. Experimental results demonstrate that PGA does not easily get stuck at a local optimum,and that it'is fast in convergence, and is simple to implment.

关 键 词：改进自适应遗传算法 双变异率 群体多样性 优化计算 测试函数 概率模型 

分 类 号：O242.23[理学—计算数学]