Smart Intentional Islanding And Random Topology Generation Of Distribution System

Connecting distributed generation to the grid and improving the reliability of power supply,especially ensuring continuous power supply for important loads are the critical issues for building efficient smart grid in our country.For this reason,the smart intentional islanding in distribution grid under fault conditions and power grid random topology generation problem is studied in this paper.Smart intentional islanding is implemented by the distribution fast simulation and modeling(DFSM)in each distribution operation center or intelligent network agent.Under this background,we introduce the technical demands,the main design ideas and the basic function architecture of DFSM,and explain the research position of island partition in the architecture.The main innovations in this paper are listed as below.Firstly,for the study on the method of island partition,a new random graph generation algorithm is proposed to effectively model the network topology.The algorithm guarantees the connectivity of the random graph through generating a spanning tree,and is able to control the number of edges in the random graph precisely.No matter how much the average vertex degree is,whether sparse or not,random graphs can be quickly generated to satisfy the requirements.At the same time,in order to simulate the power grid topology under fault conditions,the algorithm can generate the random graph with a prescribed number of connected components.Through an experimental study on several realistic power grid topologies,it is proved that the algorithm can quickly generate a large number of random graphs with realistic power grid topology characteristics.The proposed algorithm can provide a lot of random test cases for the future work.Secondly,a mathematical model of the 1-neighbour knapsack problem(1-NKP)is established,and a fast approximate algorithm for solving the 1-NKP is proposed,which called prospective greedy algorithm(PGA).The algorithm effectively overcomes the blindness of single step selection of the existing conventional greedy algorithm.A large number of experiments show that the proposed algorithm is preferable to the conventional greedy algorithm,and it is more effective to solve the 1-NKP.The mathematical model of the 1-NKP and the algorithm for solving the problem has laid a solid theoretical foundation for modeling and solving the island partition problem.Finally,based on the fact that tie switches are often neglected in the intentional islanding,a 1-NKP model for island partition is established.The model takes tie switches into account,so as to avoid islanding scheme being affected by the power grid topology before the fault happened.From the point of view of either distribution system operation or graph theory,the model is superior to the existing tree model of the distribution system.On the basis of the proposed model,island partition strategies with single DG and with multi-DG are developed respectively.And the PGA algorithm is implemented to solve the problem.The correctness of the proposed 1-NKP model for island partition,and the effectiveness of PGA algorithm is demonstrated by the application in the PG&E 69-node test case and the random generated power grid cases.