动态复杂网络的同步拓扑演化

Topological evolution on synchronization of dynamic complex networks

在定性讨论复杂网络模型的同步性能后,仿真分析了较大网络尺寸的情况,通过数据分析比对、网络拓扑的可视化和拓扑的模拟退火演化,寻找到了同步优化的一定规律,即度分布和平均距离均匀化、集中化;簇系数的适当调节对同步性能影响不大,但能减少网络连接等。结合未来智能电力网发展的实际情况,制定了同步稳定性优化策略,并在美国西部电力网数据上进行实验,探索从拓扑角度优化实际网络的应用价值,满足实时性、稳定性、分布式等要求。通过特征值比这一评价指标的对比证明优化有效。

After qualitative discussion of the synchronization performance in complex network models, simulation analysis of the networks with relatively larger size was presented. Through data analysis, network topology visualization, and topology evolution with simulated annealing algorithm, some rules of synchronization optimization were found, that is, making the degree distribution and average distance uniform and centralized, and proper clustering coefficient can reduce network connection without influencing synchronization. Considering the situation of future power grid, optimization strategies for the stability of synchronization were developed and tested on the data of the actual power grid, exploring the application value of optimizing practical networks from the angle of topology and satisfying the requirement of real-time quality, stability and distribution. The optimization is proved to be effective.