Study On Multi-objective Reconfiguration Of Distribution Network With Distributed Generations Considering Network Survivability

With the rapid development of the national economy, the economic and reliable operation of distribution network has become an important part of the power grid operation. The traditional optimization of distribution network reconfiguration is to reduce the network loss of distribution network, load balancing and improve the system reliability from the perspective of power equipment,very few to optimizate the network topology. Meanwhile, with the rapid development of distributed power generation technology, more and more distributed generations(DG) incorporated into distribution network, especially some DGs which their output are random and intermittent, make the traditional distribution network reconfiguration methods not suitable for distribution network containing DG. Therefore, it is important to consider the network topology optimization and the random output of distributed generation in the study of distribution network reconfiguration.In this paper, we firstly divide the DGs into four node types, such as PQ, PV, PI and PQ(V), and build power flow model for each node type.After analysing traditional power flow calculation method, a improved forward /backward sweep method which can deal with different types of DGs is proposed in the paper, based on the built power flow model of DGs and the node equivalent injected power model. The feasibility of the algorithm is validated by IEEE33 example.Secondly, some relative concepts are introduced in the distribution network network topoligy’s optimization.And from the point of natural connection between network nodes,a new objective function, network survivability index, for optimizing the network topology has been proposed,based on the equivalent shortest path. And, using a simple example introduces the calculation process of network survivability index.Thirdly, considering the randomness and intermittent of the output, the traditional distribution network methods have limitation. In wind power, for example, building up wind power output random variables, and their corresponding probability based on wind speed probability density function and wind power output model. And then,based on related theory of chance constrained programming,taking the switch state as optimization variables, the output of WPG as random variables, the minimum network loss higher than confidence level as another objective function,a chance constrained programming based distribution network multi-objective reconfiguration model with DG is built. Inorder to solve the distribution network reconfiguration model, the basic ant colony algorithm is improved by using the spanning tree search strategy, the dynamic path selection strategy and the multi- objective pheromone update strategy.Finally, the feasibility and effectiveness of the proposed model of distribution network reconfiguration are verified by using the IEEE33 and PG&E 69 nodes system as the test system.