| Smart distribution grid has greatly improved the reliability of power supply, however it is still possible for large area blackouts to take place under circunstances of certain irresistible external force or extreme conditions(such as ice damage,typhoon, tsunami,etc).The traditional distribution network basically applies all types of self-powered restore power to supply critical load. smart distribution network could supply the critical load or related area through micro-grid island operation due to the existence micro-grid, distributed generation(DG) and energy storage devices.This paper firstly introduces the research status of black start theory, making the distribution network’s local self-recovery strategy at early stage the core issue of smart distribution network black start research by combining intelligent optimization algorithm. Theories and algorithms concerning distribution network’s black start strategy are analyzed and researched in order to restore important load power supply as much as possible and provide theoretical and decision support for dispatchers. The main work of this paper includes the following aspects:(1) DG’s characteristics and classification are analyzed. Black start DG’s pros and cons are assessed, and the optimization study is carried through fuzzy hierarchical model. The paper discusses the factors that affect distributed power-up and establishes a fuzzy hierarchical model.The evaluation matrix is built By using the concept of triangular fuzzy numb-er,considering the indexs of which the weight value is judged by means of trapezoidal fuzzy number.The final boot sequence is optimized according to the ordination of integrated fuzzy effectiveness of each unit.(2) To make the island steadily operated, a new strategy for island partitioning, in which the operation characteristics of DG is considered, is proposed. In the proposed strategy, as many as possible key load, to which the power supply should be ensured, are taken as the objective of island partitioning, and combining classical depth-first search algorithm with binary particle swarm algorithm possessing combination and mutation the island partitioning is performed. The particles expressing partitioning scheme are updated and adjusted by binary particle swarm algorithm with combination and mutation; for new particles generated by each iteration the connectivity of single island power is verified by the depth-first search algorithm; for the island, which does not meet the requirement of connectivity, the encoding scheme is adjusted, finally the optimal scheme is found out. The effectiveness of the proposed method is verified by simulation results of IEEE69-bus system.(3) A DG-based local self-recovery scheme for distribution network at post-blackout early stage of power grid restoration is analyzed. The starting sequence of DG after the partitioning and the optimization of the restoration path are discussed to ensure that the power supply for as many as possible loads can be recovered within as short time period as possible. In the proposed model, in which the weight of minimum path and restoration of power supply for maximum essential loads are taken as objectives and the recovered power supply for loads not exceeding the generating capacity of DG is taken as one of constraints, is solved by the modified particle swarm optimization (MPSO). Simulation results of IEEE33-bus system show that the proposed method is effective.In this paper, in order to restore the power supply to the important loads of the smart distribution network under some extreme conditions, micro-grid and DG are used to divide the distribution network into several small-scale systems. Then the power supply for each area’s loads will be restored orderly through blackstart. At last, grid can be connected after the recovery of transmission line. |
PDF Full Text Request