Research On Optimization And Scheduling Of Active Distribution Network With Load Regulation

The operation management of traditional distribution network is to plan and predict the recent electricity production by collecting data from past years,so as to meet users’ needs.However,the rapid development of clean energy has increased the energy consumption of the source side of the grid,which has changed the traditional operation mode of the distribution network.At the same time,the popularity of the new controllable load also poses a challenge to the active management technology and user satisfaction of the demand side of distribution network.Under this environment,it is necessary to study load forecasting and multi-objective optimization dispatching of active distribution network with distributed energy.In this paper,the research status of medium and long term load forecasting and active distribution network optimization scheduling is introduced,and the characteristics of "source,storage and charge" unit of active distribution network are analyzed.Based on the grey system theory as the model foundation of mid long term load forecasting,the classical grey model is introduced and analyzed,and a multivariate MGM(1,n,r)prediction model suitable for active distribution network is built.Through the analysis of multivariable grey model,the feasibility of medium and long term load forecasting under the environment of active distribution network is demonstrated.The optimization characteristics of genetic algorithm GA parameters are analyzed.Then based on the GA optimization of MGM(1,n,r)medium and long term load forecasting example analysis,the prediction process is derived,and the error reasoning is carried out.Through the historical data of influencing factors of load development in an area of Xinjiang in recent ten years,the power supply load and various analysis,respectively using single variable GM(1,1),MGM(1,n)and GA optimization MGM(1,n,r)model in load forecasting,the results demonstrated that the proposed improved load the prediction model is better than the traditional grey forecasting load model,and is more suitable for multi factors of active distribution network load environment.Finally,by considering the load end peak,a multi-objective optimization scheduling model of active distribution network,select the extended IEEE33 node distribution system verifies the model to calculate the optimal solution of Pareto,analysis of each unit before and after optimization of running state,demonstrate the practicality of this economic multi-objective optimization model.