Coordinated Optimization Dispatch Method Of Distribution Networks With Microgrids Under Uncertainty

In order to solve problems of energy shortage and climate warming,and to achieve the long-term goal of carbon neutrality,it is necessary to greatly increase the proportion of renewable energy generation from wind and solar energy.As an efficient carrier of distributed renewable energy connected to the distribution system,microgrid has received extensive attention from academic and engineering circles.In order to solve the influence of uncertain factors in microgrid operation on system stability and economy,and to give play to the friendly interaction and auxiliary services of the microgrid to the distribution system,this thesis focuses on the coordinated optimization dispatch method of distribution networks with microgrids under uncertainty.The main work is as follows:1)A robust optimization method of microgrid considering multiple correlations of wind power is proposed for the day-ahead economic dispatch of grid-connected microgrid.Based on the multivariate normal copula theory,considering the temporal correlation of the wind power forecast error and the conditional correlation with the forecast value,a data-driven multi-ellipsoidal uncertainty set modeling method is proposed,and an improved multi-ellipsoidal uncertainty set is established by proposing a comprehensive index to optimize the dimension of ellipsoid.The uncertainty set is applied to the robust optimization model of microgrid,and it is transformed into a solvable form according to the strong duality theory and multiple linearization techniques.Numbers of cases verify that the method has higher uncertainty modeling accuracy and can effectively reduce the conservativeness of traditional robust optimization methods.At the same time,in the case of limited training samples,it has lower unbalanced power and total operation cost than other optimization methods.2)Aiming at the short-term line capacity shortage problem of distribution network in the heating season under the background of "coal to electricity" in rural areas of northern China,a two-layer coordinated optimization dispatch method of distribution networks with microgrids based on multi-agent game is proposed.The microgrid and heat pump clusters obtain the hourly optimal power plan through hourly rolling optimization within the day.The distribution network verifies whether the plan will lead to tie line power exceeding the limit,and issues auxiliary service requirements.The auxiliary service mechanism based on the two-layer coordinated optimization model enables the distribution network to minimize operation cost,while the microgrid and heat pump clusters can maximize the benefits of auxiliary services through game competition.The decentralized Wo LF-PHC(Win or Learn Fast Policy Hill-climbing)algorithm is used to solve the multi-agent matrix game problem,which can quickly realize the game equilibrium under limited information interaction and protect the privacy of the participants.The simulation verification of the algorithm is completed based on an actual project case in China,and the results show that this method can effectively solve the problem of insufficient line capacity of distribution network,and enable microgrid operators and heat pump cluster agents to obtain auxiliary service benefits,which achieves a win-win situation for multiple entities.This study has a certain reference value for the construction of distribution network auxiliary service market in the future.