基于复杂网络理论的换热网络节点重要性研究

发布时间：2018-03-01 18:53

本文关键词： 换热网络复杂网络节点中心性 PageRank算法旁路　出处：《青岛科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着化工工艺过程的大型化和复杂化,其子系统换热网络也变得相对复杂,对其控制也更加困难,换热网络中换热器的换热效能发生改变,其干扰传递会导致与之相关联的换热器换热效能改变,由于系统耦合严重,进而影响到整个换热网络的操作性和稳定性。为此,对换热网络中的换热器节点进行重要性研究,从而针对特定的换热器进行必要的监测和设置旁路,具有重要意义。首先,在文献综述的基础上,比较各个建模方法,选择基于干扰传递构建换热网络的复杂网络模型,并对网络的复杂性、小世界性和度分布特性等进行描述和判断。其次,为了全面综合地评价该网络模型,先从网络的拓扑结构出发,不考虑网络的方向性,定量计算网络每个节点的点度中心性、中间中心性、接近中心性和特征向量中心性,由于各个指标的侧重点不同,提出了基于多属性决策算法处理相关指标的方法,得到基于拓扑结构的换热网络节点重要性评价数据;在此基础上,进而考虑网络的方向性,以PageRank算法对网络的节点重要性进行评价,综合两种不同的算法结果得出最终结论。最后,以复杂网络可控性研究理论为基础,分析网络的邻接矩阵,结合节点重要性排序和网络严格可控的必要条件,确定一组网络控制所需要的最优驱动节点组,将驱动节点组所在的换热器进行旁路设计。通过案例分析,证明了本方法的有效性,该方法能够有效地确定换热网络节点重要性,并为换热网络的旁路优化设计提供理论支持。
[Abstract]:As the chemical process becomes larger and more complicated, the heat exchange network of its subsystems becomes more complex, and it is more difficult to control it. The heat transfer efficiency of the heat exchanger in the heat exchanger network is changed. Its interference transfer will result in the change of heat exchanger heat transfer efficiency associated with it. Because of the serious coupling of the system, the operation and stability of the whole heat exchanger network will be affected. Therefore, the importance of the heat exchanger node in the heat exchanger network is studied. It is of great significance to monitor and set up bypass channels for specific heat exchangers. Firstly, on the basis of literature review, various modeling methods are compared, and complex network models based on interference transfer are selected to construct heat exchanger networks. The complexity of the network, the characteristics of the small world and degree distribution are described and judged. Secondly, in order to evaluate the network model comprehensively and synthetically, the topological structure of the network is considered first, and the directionality of the network is not considered. Quantificationally calculate the vertex centrality, middle centrality, proximity centrality and eigenvector centrality of each node in the network. Because the emphasis of each index is different, a method based on multi-attribute decision making algorithm to deal with related indexes is proposed. The node importance evaluation data based on topology structure are obtained, and then, considering the directivity of the network, the importance of nodes is evaluated by PageRank algorithm. Finally, based on the theory of controllability of complex networks, the adjacency matrix of networks is analyzed, combined with the order of importance of nodes and the necessary conditions of strict controllability of networks. The optimal driving node group is determined for network control, and the heat exchanger in which the driving node group is located is designed by bypass. The effectiveness of this method is proved by case study, and the importance of heat transfer network node can be effectively determined by this method. It also provides theoretical support for bypass optimization design of heat transfer network.
【学位授予单位】：青岛科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ051.5

【参考文献】