Zeng Fang, Li Xiang, Li Ke
Faculty of Engineering, University of Georgia, Athens, Georgia 30605, USA.
Department of Computer Science, University of Georgia, Athens, Georgia 30605, USA.
Chaos. 2017 Feb;27(2):023105. doi: 10.1063/1.4975762.
The complex topology and adaptive behavior of infrastructure systems are driven by both self-organization of the demand and rigid engineering solutions. Therefore, engineering complex systems requires a method balancing holism and reductionism. To model the growth of water distribution networks, a complex network model was developed following the combination of local optimization rules and engineering considerations. The demand node generation is dynamic and follows the scaling law of urban growth. The proposed model can generate a water distribution network (WDN) similar to reported real-world WDNs on some structural properties. Comparison with different modeling approaches indicates that a realistic demand node distribution and co-evolvement of demand node and network are important for the simulation of real complex networks. The simulation results indicate that the efficiency of water distribution networks is exponentially affected by the urban growth pattern. On the contrary, the improvement of efficiency by engineering optimization is limited and relatively insignificant. The redundancy and robustness, on another aspect, can be significantly improved through engineering methods.
基础设施系统的复杂拓扑结构和自适应行为是由需求的自组织和严格的工程解决方案共同驱动的。因此，工程复杂系统需要一种平衡整体论和还原论的方法。为了模拟配水管网的增长，在结合局部优化规则和工程考虑因素的基础上，开发了一种复杂网络模型。需求节点的生成是动态的，并遵循城市增长的标度律。所提出的模型在某些结构特性上能够生成与已报道的实际配水管网相似的配水管网。与不同建模方法的比较表明，现实的需求节点分布以及需求节点与网络的共同演化对于真实复杂网络的模拟至关重要。模拟结果表明，配水管网的效率受到城市增长模式的指数影响。相反，通过工程优化提高效率的作用有限且相对不显著。另一方面，冗余性和鲁棒性可以通过工程方法得到显著提高。
