Optimal Configuration Of Distributed Generation In Active Distribution Network Based On Bi-level Programming Theory

With the increasing shortage of traditional fossil energy sources and the increasing demand for electric energy,the high proportion of distributed generation represented by wind power and photovoltaics has become the development trend of the distribution network.In order to cope with a series of problems after distributed generation connected to the grid,the traditional distribution network is gradually transforming into an active distribution network with active management measures.Starting from the introduction of the active distribution network,this thesis analyzes the impact of distributed generation after being connected to the distribution network.Then,in order to consider the uncertainty of distributed generation and load,the probability model that wind speed obeys Weibull distribution,light intensity obeys Beta distribution and the probability model that load obeys normal distribution are established respectively.The scene generation technology and the fast forward selection scene reduction technology are used to generate a typical scene model.Based on the active distribution network,this thesis analyzes and studies the optimal configuration of distributed generation.First,without considering the uncertainty of wind power,an optimal configuration model with the goal of minimizing active power loss is established.Taking the IEEE33-node system and IEEE69-node system as examples,genetic algorithm,particle swarm optimization algorithm,grey wolf optimization algorithm and Archimedes optimization algorithm are used to verify and analyze the established optimization.And the simulation models are built on the ETAP platform to verify the configuration schemes.The results show that the configuration schemes obtained by the grey wolf algorithm and the Archimedes optimization algorithm are optimal.Then,considering the uncertainty of wind power,a bi-level planning model of distributed power generation for active distribution network is established.The upper-level model considers the minimum investment cost of distributed generation as the goal,and the lower-level takes minimizing the system network loss as the optimization target.For the solution of the upper-level problem,an improved grey wolf optimization algorithm with the stronger global optimization ability and higher solving efficiency is proposed.When solving the lower-level model,convex relaxation and linearization techniques are used to transform the lower-level model into a Second-order cone programming problem.Finally,take the IEEE33-node system and IEEE69-node system as examples to verify the bi-level programming model established in this thesis.Meanwhile,the planning schemes are further simulated on the ETAP platform.The results show that the model established in this thesis can obtain a reasonable planning scheme.