Research Of Network Reconfiguration Strategy For Yunnan Power Grid

As an important route of power transmission from west to east, Yunnan power gridhas developed rapidly in recent years. With the gradual increasing scale and theescalating interconnection links of power grid, operation adjustment and control becomemore complex, and the potential risk of blackout is also increasing. The recovery processof system after blackout lasts a long time, and has many contents. Each stage hasintimate connections and interacts with each other. Successful implementation of blackstart operation must be based on the resume program which coincides with the actualrunning state of the system. Thus, the generation principles and methods of the objectivenetwork in initial time of black-start restoration is presented in this paper according to thenetwork structure and the actual condition after blackout, which reflects the timelinessand operability. Then practical system partitioning scheme and network reconfigurationstrategy are developed to provide decision support for the restoration of Yunnan powergrid after blackout on the basis of power grid optimization principle.With deep analysis of Yunnan power grid in the grid structure, the principles andpower supply of black-start, combining the characteristics of network structure, themethod to generate the objective network in initial time of black-start restoration is putforward, which searches topology database quickly in a way of "network layer+plant-station layer ", and the scheme checkouts are also completed in this paper. All theanalysis is based on the system state identification of running section obtained from theOCS system of Yunnan power grid. The effectiveness of the proposed method isvalidated by the mid-west area restoration of Yunnan Power Grid.According to the deficiency of the existing black-start partitioning scheme ofYunnan power grid that is primarily based on geography division and lack of scientifictheory, spectral clustering algorithm is used to deeply excavate the community structureof the grid to achieve partitioning scheme. First, calculate the Laplace matrix of thesimplified grid and obtain the clustering samples formed by the first and second smallnon-trivial eigenvectors. Then the initial clustering centers are determined combined withthe number and distribution of black-start power supply. At last, the preliminarypartitioning scheme is realized by using the k-means algorithm to cluster grid nodes. Onthis basis, according to Yunnan power grid black-start preplans and the actual conditionafter blackout, the further reasonable adjustment method of system partitioning is made,and then final partitioning scheme is achieved which accords with the requirements ofpractical application in Yunnan power grid.For the uncertainty of operating time and recovery reliability when lines put intooperation during network reconstruction after blackout, a fuzzy chance constrainedmodel of optimization of network reconfiguration is proposed. In this paper, triangular fuzzy variables are used to represent line’s operating time and recovery reliability, andthe recovery reliability index is defined to evaluate the objective network. The target is toensure that the possibility of unit’s restore in its start-up time requirements is not lessthan a certain level of confidence, based on which obtain the optimized network thatreaches the highest reliability in a shortest possible time. The fuzzy simulated based crossparticle swarm optimization is used to optimize the node restoration sequence, and theDijkstra algorithm is combined to choose the restoration path. At last, the effectiveness ofthe proposed method is validated by the IEEE30-bus system and part of Yunnan powergrid.