Research Of Optimal Dispatch Method For Distribution Network With Photovoltaic Microgrid

With the rapid development of the national economy,the demand for energy is increasing.The exploitation and utilization of renewable energy sources are very important for such two problems,which has attracted much attention.The renewable distributed grid reaches large power grid in a friend manner through microgrid,which can improve the stability and reliability of power supply system.This thesis studies the generation output rate forecasting technique of photovoltaic microgrid and the optimal dispatch of distribution network with photovoltaic microgrid,which mainly include:1)The establishment of mathematical model for photovoltaic microgrid: the mathematical model for the power units,energy storage units,controllable distributed generation units and load units in the photovoltaic microgrid.This thesis will mainly states the principle of photovoltaic generation units and its output rate characteristics,which provides the basis for the forecasting of photovoltaic microgrid generation output rate and the optimal dispatch of distribution network with photovoltaic microgrid.2)The study of forecasting technique of photovoltaic microgrid generation output rate.Aiming at the inexact forecasting of photovoltaic microgrid output rate,one synthesized short-term forecasting method of the photovoltaic microgrid output rate has been put forward.The method takes full use of the Neural Network Algorithm that can solve the nonlinear problem,combining with the advantages of direct and indirect forecasting of photovoltaic microgrid output rate.The date of temperature,solar radiation intensity and historical power output rate(close to the data of the forecasting day)issued by weather bureau is taken as a sample to train neural network.The ground-based cloud image is adopted to analyze,process and forecast the solar irradiation attenuation caused by the cloud cover.The data of temperature,solar irradiation attenuation and cloud attenuation acts as input to gain preliminary forecasting result.And this result will be further amended through the law which is drawn by the analysis of the historical power data fluctuation,and this will improve the forecasting accuracy of photovoltaic power output rate.Based on MATLAB,the simulation analysis is carried out to verify the validity of improving forecasting accuracy through the comprehensive method in this thesis.3)Study of the optimal dispatch of distribution network with Photovoltaic microgrid,the reactive and active the optimal dispatch of distribution network with Photovoltaic microgrid.In this thesis,the minimum cost maximum flow model is taken to carry out an active optimal dispatch,and to turn the nonlinear problem in active optimization into a linear one.Flower Pollination Algorithm Based on Gravity Search Mechanism is employed to carry out the reactive optimization dispatch of distribution network with microgrid.The combination of Gravity Search Mechanism and Flower Pollination Algorithm,individual Flower Pollination Levy mechanism in group and interindividual universal gravitation is taken to update the flower individual location,improve the global optimization ability,accelerate convergence and effectively find out the locally optimal solution.Simulation experiment shows that the model in this thesis can effectively reduce operational costs,lower internet fee and improve voltage qualified rate,compared with the optimal dispatch method of active distribution network which is based on day-head scheduling plan cure.