Multi-Attribute Partitioning of Power Networks Based on Electrical Distance

摘要: Identifying coherent sub-graphs in networks is important many applications. In power systems, large systems are divided into areas and zones to aid planning control But not every partitioning equally good for all applications; different applications have goals, or attributes, against which solutions should be evaluated. This paper presents a hybrid method that combines conventional graph algorithm with an evolutionary partition network optimize multi-attribute objective function based on electrical distances, cluster sizes, the number of clusters, connectedness. Results IEEE RTS-96 show clusters produced by this can used identify buses dynamically voltage angles, without need dynamic simulation. Application 118-bus 2383-bus case indicates when well partitioned zones, intra-zone transactions less impact flows outside zone; i.e., reduces loop flows. property particularly useful system where ensuring deliverability important, such as transmission determination synchronous reserve zones.