集群行为中的复杂网络结构合并优化方法

发布时间：2019-04-30 15:29

【摘要】：集群行为中的人际互动关系可以通过复杂网络结构予以表示,以特定结构特征为目标的结构合并优化问题对于研究集群行为具有很好的应用价值。本文通过数理模型分析,以最小平均路径长度为优化目标,以不同连边策略为变量构建了复杂网络合并优化模型,提出了启发式遗传算法的网络结构合并优化SCOA算法。最后,针对不同复杂网络模型进行实验与仿真,讨论集群行为群体合并的一般策略与相关特征。研究发现,不同的连边规则对网络结构合并性能有显著影响,网络结构合并优化问题客观存在;本文提出的SCOA算法优于现有的同配性连边规则、异配性连边规则与随机连边规则;其所得的连边规则是一个包含同配连边规则与异配连边规则的混合连边规则。本文研究为集群行为中解决群体隔离问题提供了新的分析思路与方法。
[Abstract]:The interpersonal interaction in cluster behavior can be represented by complex network structure. The optimization problem of structural merging with specific structural characteristics is of great application value to the study of cluster behavior. In this paper, a complex network merging optimization model is constructed by mathematical model analysis, taking the minimum average path length as the optimization objective and different connecting strategies as variables, and a heuristic genetic algorithm (GA) SCOA algorithm for network structure merging optimization is proposed. Finally, experiments and simulations are carried out for different complex network models, and the general strategies and related characteristics of cluster behavior group merging are discussed. It is found that different linking rules have a significant effect on the performance of network structure merging, and the optimization problem of network structure merging exists objectively. The SCOA algorithm proposed in this paper is superior to the existing homographies, heterogamy and random edge-joining rules, and the resulting joining rules are a hybrid joining rule which contains the same-matching edge rule and the heterogrimalized edge-joining rule, and the result of the proposed algorithm is better than that of the existing homographies, heterogamy and random edge-joining rules. This paper provides a new approach to solve the problem of colony isolation in cluster behavior.
【作者单位】：长安大学经济与管理学院;西安交通大学公共管理与复杂性科学研究中心;西安交通大学公共政策与管理学院;
【基金】：国家自然科学基金资助项目(71071128) 国家社会科学基金重点项目(12AZD110) 中央高校基本科研业务费专项资金项目(西安交通大学基本科研业务费青年教师跟踪支持计划) 国家社会基金资助项目(17BJY139)
【分类号】：C912.3

【相似文献】