Research Of Some Key Techniques Of Distribution Network With Distributed Generations

With the gradually practical use of smart grid in our country,distributed generation(DG) with its good compatibility is gradually beingapplied to the distribution network. When connected to the distributionsystem, DG could provide a series of advantages such as regulating thedifference between peak power, energy storage, saving investment andproviding uninterrupted power to the critical load. However, DG changedthe original grid power flow, line current, voltage, and system losses. Forrandom fluctuations type DG, such as doubly-fed wind generation, it couldcause node voltage fluctuations and voltage flicker phenomenon. Therefore,it’s very necessary to deal with the distribution network protection with DGs,reactive power optimization and other key technology issues.Firstly, according to the characteristics of DG grid, the different formsof influence of DG on the distribution system are analyzed in detail. Then,based on PSCAD model, the control structure of a new FCL topology isstudied, and the effectiveness of the fault current limiter in reducing theoccurrence of various types of faults is verified through examples of a realnetwork model in this paper. Secondly, combining with the basic principlesof traditional time inverse over-current protection, a new DG locating andsizing method and protection strategy based on FCL is studied. Thirdly,combined with the characteristics of randomly fluctuating DG, a kind ofcontrol strategy using dynamic reactive power compensation of SVG is putforward to maintain the stability of randomly fluctuating nodes’ voltage.Based on a detailed analysis of determing the capacity of SVG, a newreactive power optimization strategy of distribution system with DG is proposed. SVG and output adjustable reactive power are employed tofine-tune, and the static compensator device are used to comprehensivelyregulate the reactive power in power system. Finally, a reactive poweroptimization strategy based on fuzzy control theory and particle swarmoptimization algorithm are presented in this paper, with examples of IEEE9distribution network, which verifying the rapid usability and effectivenessof the optimization strategy as well as the improved algorithm.