Optimal Configuration Of Distributed Power Generation Considering Demand Response And Timing Characteristics

With the increasing energy crisis and environmental pollution and the opening of the distributed generation market,the distributed power generation(distributed generation,DG)led by wind and solar power is booming.Because the distributed power generation has obvious timing and volatility a large number of access to the distribution network will not only generate energy consumption problems,but also make the operation of the power system more complex.In order to coordinate the supply-demand relationship between new power sources such as DG and between DG and load,and solve the problems of voltage imbalance,wind and light abandonment caused by DG high penetration access,active distribution network(ADN)with active management and operation capability will become the main development form of distribution network in the future.At the same time,Demand Response(DR)load is the main interactive resource of DG under the ADN mode.It can adjust the demand of power load,optimize the load power consumption mode,and play a vital role in load peak shaving and valley filling,and improve the interaction level between the grid and users.Therefore,reasonable configuration of distributed power supply and use of demand side management are of great practical significance to improve the safe,stable and economic operation of ADN.In this paper,the optimal configuration of distributed generation in active distribution network considering the timing characteristics of wind and load is studied.The main work is as follows:Firstly,based on the time series characteristics between DG generation and demand side,a probabilistic statistical model of wind power output and customer side time series demand considering the time series characteristics is constructed;Using Latin hypercube sampling and Nataf transform,the specific probability distribution characteristics of each time period are classified,and the all-hour wind-light-charge output scene is constructed considering the temporal correlation.In order to improve the diversity and rationality of clustering scenarios and reduce the difficulty of model operation and solution,this paper introduces Davies-Bouldin Index(DBI)and K-means algorithm to cluster and reduce full-hour scenarios,so as to determine the number of typical scenarios meeting the effectiveness index.Secondly,a two-level optimal configuration model of active distribution network considering the time series characteristics of wind power and load is established.The upper level takes the optimal annual comprehensive economic cost of the system as the goal,and the decision variables are the location and capacity of wind power units connected to the distribution network;The lower level takes the optimal operation voltage level of the system at each time period as the goal,and the decision variable is the operation cost under the active management measures(AM)of the distribution network;By comprehensively considering the timing output of landscape unit and load side timing response,the economic planning of distributed power supply of active distribution network is realized.According to the established model,the algorithm introduces the strategy of roulette selection and elite retention,and at the same time,the crossover and mutation probability are appropriately improved.The improved adaptive genetic algorithm(IAGA)is obtained.An IEEE 33-bus system example shows that the proposed method and model not only ensure the diversity and rationality of scenario selection,but also take into account the characteristics of wind turbine output and load timing fluctuations,improve the system operation level and reduce the system operation cost.Finally,in order to make full use of the resources on both sides of source-load,based on the time-series characteristics of wind-light-load,the DR mechanism is integrated into the DG grid connection planning,and a distributed generation grid connection planning model taking into account the demand-side response is established.The lower level is the short-term operation optimization under the DR mechanism,and the upper level determines the location and capacity of DG installation based on the lower level’s comprehensive cost.The above model is simulated by using the GUROBI solver on the MATLAB platform,which verifies the rationality and effectiveness of considering DR mechanism when DG is connected to the grid.