摘要
为解决大规模复杂优化问题,针对遗传算法的并行化原理和常用运行平台进行分析,提出了并行遗传算法新的应用平台———Internet,讨论了基于Internet的并行遗传算法(Internet basedparallelgeneticalgorithms,IPGA)实现中的关键问题,并给出其单向环拓扑的具体实现。实验表明,IPGA可显著节约寻优时间,提高寻优质量,并且能够充分利用互联网中闲置的计算机资源,节约运行成本,有助于解决巨量优化问题。
In order to solve the massive complicated optimization problems, the parallelization principle and the current application platform of genetic algorithm are analyzed, and a new application environment, Internet, is presented. Then the key problems of IPGA (Internet-based parallel genetic algorithm) are discussed. In the end, the ring topology is also implemented. Experimental results demonstrate that IPGA can not only evidently save the optimization time but also largely improve the optimization quality. At the same time, the algorithm can make full use of the idle computers on Internet and reduce the running cost, providing an effective solution to massive optimization problems.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2004年第8期1102-1106,共5页
Systems Engineering and Electronics
