Reactive Power Optimization Of Distribution Network With Distributed Generation Based On Scenario Analysis

Reactive power optimization of electrical power system is a complicated nonlinear optimization problem, which has the characteristics of multiple-constraint, multi-objective and control variables discretization. Reactive power optimization can improve the system voltage level, reduce the network loss and ensure safe and stable operation of power system. Renewable energy’s access to distribution network can relieve the energy crisis, promote the development of energy conservation and emission reduction, but on the other hand, it would change the power flow of the distribution network, and bring new challenges to planning, operation and control of the distribution network, restricting the development of distributed generation. At the same time, some traditional method for reactive power optimization will fail because of not considering the influence of distributed generation. In this paper, the research is on reactive power optimization of distribution network containing distributed generation, the main work is as follows:(1) In this paper, the mathematical model for reactive power optimization of distribution network is constructed, which is with objective function of reducing the network loss, and in order to reduce system voltage deviation, the penalty function to punish the node voltage is introduced to objective function. Meanwhile considering the distributed generation has the ability of reactive power regulation, the reactive power output of distributed generation and the number of switched capacitor group are usually set as control variables; voltage amplitude of load node is set as state variable. In this model, the reactive power optimization of distribution network with distributed generation is realized through the optimization of reactive power output of distributed generation and reactive power compensation of capacitor group.(2) This paper mainly researches on the access of wind and photovoltaic power generation into distribution network. The output power of wind turbine usually changes with wind speed, the output power of photovoltaic power generation is mainly influenced by solar illumination intensity, which both have randomness characteristic. Scenario analysis method from the view point of probability is adopted to solve the output of intermittent of distributed generation. According to the probability density functions of output power of wind and photovoltaic power generation, the combination scenes composed of definite factors are built to eliminate the uncertainty of output power of distributed generation.(3) In this paper, the hybrid algorithm of Metropolis-Hasting and improved particle swarm optimization (PSO) is proposed. This algorithm combines the two algorithms through making Metropolis-Hasting sampling for particles, the specific step is to make the particles search nearby according to normal distribution, and then determine the next generation of particles through comparing the size of the accepting probability and a random number. Based on the advantages of PSO, such as simplify, fast convergence speed and less parameters, this hybrid algorithm can effectively overcome the shortcoming of local convergence, enhance population diversity and obtain high convergence precision and speed.(4) At last, a simulation of IEEE 33-bus network can verify the effectiveness of the reactive power optimization model and the hybrid algorithm of Metropolis-Hasting and PSO, which can realize the reactive power distribution, improve the system voltage level and enhance the operation stability of power system.