基于混合遗传算法的园区网络优化设计研究-情报学专业论文.docxVIP

摘要摘要 摘要 摘要 本文主要以园区网络的优化设计为目标，综合运用图论、集合论、运筹学、 数理逻辑、系统工程等理论和技术，分析传统网络优化设计方法和标准遗传算法 的不足之处，进而研究了混合遗传算法在园区网络优化设计中的应用。 重点分析了园区网络设计的各方面因素及网络优化设计的传统方法。传统方 法不具备全局寻优的能力，进一步分析标准遗传算法理论，浚算法在解决优化问 题时缺乏局部寻优的能力，针对上述不足提出了一种混合遗传算法 ．结合饱和割 集算法的遗传算法。 建立了使用该算法进行园区网络优化设计的模型，对原有遗传算法的编码机 制、种群初始化算子、选择算子、交叉算子、变异算子进行了改进，并在变异算 子中引进饱和割集算法。最后，通过具体实例验证该模型效果优于传统方法。 本研究课题得到研究项目汉中市电子政务系统集成(00235P01)的支持。本 论文是该研究项目的理论总结和提高，该研究项目被评为陕西省电子政务系统建 设示范工程。 关键字：园区网络优化设计混合遗传算法饱和割集算法 网络设计方法 absttactAbstract absttact Abstract This paper mainly deals witll the optimization design of campus networks．The graph theory,the set theory,the operational research theory,the logic and the system engineer are applied in this paper．The limitations of the traditional network design methods and the standard genetic algorithm are analyzed and the application of the hybrid genetic algorithm is also explored．The main points of the study in this paper are aS follows． By analysing the factors and the traditional methods of the campus network design，it summarizes these methods are lack of global search ability．It also analyses the standard genetic algorithm，and summarizes it is lack of local search ability．On the basis of these，a hybrid genetic algorithm mixing the standard genetic algorithm and the cut saturation algorithm is proposed． It sets up the model of the campus network optimization design based on the hybrid genetic algorithm．By this algorithm，the code mechanism，the population initialisation，the selection operator,the crossover operator and the mutation operator have been improved．At last，it is verified that the new method is better than the traditional ones through the practical application． This research subject is supported by the Project of Hanzhong Municipality’S Electronic Administration System．(00235P01)．The outcome of the project was awarded the demonstration project of the Shanxi’S Electronic Administration System． Tllis paper is the summary and improvement ofthe researoh． Keywords：Campus network Optimization desig