Research On Important Nodes Recognition And Transient Performance Of Power Networks

In recent years,with the mutual penetration between natural science and social science,complex networks have become a new research hotspot.Application of complex network theory,methods and ideas to study the complexity of the network era scientific issues reflect the trend of comprehensive development of science.Power network is a typical complex network,its synchronous and stable operation is the fundamental guarantee for people's normal life and work.Using the complex network theory to analyze the topology and dynamic characteristics of the power network can help us understand the structural characteristics and operation mechanism of the power grid deeply,which help us take targeted measures to prevent the occurrence of cascading failures to ensure the normal operation of the grid.Based on the complex network theory and the topology of the actual power network as well as the dynamic characteristics of the nodes,a series of important research on the important node recognition and transient performance of the power network is carried out.The main contents include the following aspects:1.Based on the topology and dynamic characteristics of the power network,we study the importance of nodes in the IEEE test grids.Considering the global and local characteristics of a network,seven evaluation indexes,such as degree centrality,closeness centrality,sub-graph,PageRank,the nodal cohesion,the reciprocal of the discontinuous diffusion time and the critical synchronous coupling strength,are respectively used to describe the importance of nodes.In the IEEE standard test networks,we use these seven indicators to identify the importance of nodes and compared the identification results of different single indicators.Studies have shown that a single indicator considers only one aspect of the network's attributes,so the importance of a node identified by a different single indicator is different,and the important node recognition has some limitation using only a single indicator.2.Based on the analytic hierarchy process,we use the multi-index synthesis method to study the important nodes in the power network.We do comprehensive research by degree centrality,closeness centrality,PageRank,the nodal cohesion,the reciprocal of the discontinuous diffusion time and the critical synchronous coupling strength,because their recognition results are better.The weights of the five indexes are determined by the analytic hierarchy process(AHP),and the comparison matrix of the five indexes is constructed by the range method.Finally,the order of the important nodes in the network is obtained under the guidance of the multi-index comprehensive evaluation idea.Then,considering the dynamic characteristics of the power network,we only remove one node at a time,and then use the Kuramoto model of the power network to simulate the average path length,relative scale and critical synchronous coupling strength of the maximum connected subnet in the network.By comparing these three parameters to verify the multi-index node comprehensive recognition effect.The results of the verification show the effectiveness of the multi-index comprehensive evaluation method.3.We study the transient performance caused by distributed power station.First,the transient duration and the maximum frequency deviation are defined to describe the transient performance.And then we studied the selection of the parameters of the distributed power station.In the case of ensuring the frequency of the distributed power station and the operating frequency of the original grid,we found that there is an optimal network phase in the distributed power station.When the distributed power station access to the power network by the best phase,it has the slightest interference to the network.We investigated the effect of the importance and degree of load nodes on the transient performance of power networks.The research shows that the distributed power station which accesses the power network through the important nodes caused a smaller interference than which accesses the power network through the non-important nodes.In many cases,the distributed power station which access to the large degree nodes induced a smaller interference than which accesses to the slight degree nodes.Finally,in the case of the same proportion of distributed power plants,we compared the transient performance of IEEE14,IEEE30and IEEE57 three standard test power network.The results show that in the case of the same proportion of distributed power stations,the more the number of the original grid nodes,the more connected the actual power station,the worse the transient performance.