Study On Coordination Strategies Of Day-ahead And Intra-day Time Scales For Optimal Dispatch In Active Distribution Network With Distributed Energy Storages

Large-scale of utilization of renewable energy have become an inevitable trend of global energy development,and active distribution network is an effective technical mean to consume renewable energy.However,renewable energy has the characteristics of randomness,volatility and intermittence.These bring new challenge to the operating control and optimal scheduling of active distribution network.But optimal scheduling strategies of traditional distribution network are difficult to coordinate and optimize the scheduling of various distributed generators.Distributed energy storages have been paid more and more attention in recent years.And distributed energy storages participating in the optimal dispatch of distribution network can effectively reduce the operating costs of distribution network,stabilize fluctuations of renewable energy generation,realize peak load shifting,etc.Therefore,it is of great significance to study optimal scheduling strategies of active distribution network with distributed energy storages.At present,optimal scheduling strategies of active distribution network with centralized energy storages have been deeply studied,but optimal scheduling strategies of active distribution network with distributed energy storages are not perfect,and the analysis of the regulation capacity of distributed energy storage is insufficient.In particular,no research has considered the impact of the schedulable ability of distributed energy storages on the optimal scheduling of active distribution network.Then most researches have focused on the day-ahead scheduling of active distribution network based on short-term prediction of renewable energy generation.With prediction time scales increase,the prediction error of renewable energy generation increases.So the short-term forecasting error of renewable energy generation is larger than the intra-day ultra-short-term forecasting error of renewable energy generation.Accordingly,the day-ahead scheduling of active distribution network is unreasonable.Therefore,how to reasonably formulate optimal scheduling strategies considering the schedulable ability of distributed energy storages is important issue in the research of the optimal scheduling of active distribution network with distributed energy storages.In view of the above problems,on the basis of fully analyzing the regulation capability of distributed energy storages,the paper studies day-ahead and intra-day time scales coordinative dispatch strategy of active distribution network considering the schedulable ability of distributed energy storages.Firstly,based on the qualitative analysis of the energy supply characteristics of distributed energy storages,a method of regulating capacity analysis of energy storage system based on fuzzy comprehensive evaluation is proposed to further quantify the regulating capacity and peak shaving characteristic of distributed energy storages.In addition,in order to improve the accuracy of optimal dispatching,a node forecasting model based on deep learning is established to improve the accuracy of renewable energy generation and load forecasting.Then a schedulable ability evaluation model of distributed energy storages is proposed.Finally,on the basis of the characteristic analysis of energy storage system and forecasting evaluation model,a day-ahead and intra-day time scales coordinative dispatch strategy of active distribution network considering the schedulable ability of distributed energy storages is proposed to improve the impact of uncertainty of renewable energy generation caused by prediction time scale on optimal dispatching of active distribution network.The main research contents of this paper include:(1)An analysis method based on fuzzy comprehensive evaluation for adjusting capacity and peak shaving characteristic of distributed energy storage is proposed.Aiming at the problem of how to fully analyze the regulating capacity of distributed energy storages,this paper firstly qualitatively analyses energy supply characteristics of common distributed energy storages from the response speed,power regulation characteristic,etc.Then an evaluation model of regulating capacity of energy storages based on fuzzy comprehensive evaluation is established to quantitatively analyze the adjustment ability of distributed energy storages,and simulation results verify the effectiveness of the proposed model in quantifying the regulating capacity of energy storages.Finally,models of distributed energy storages accessing distribution network are constructed,and the peak shaving characteristic of distributed energy storages is analyzed.This part is the research basis of this paper.(2)A method based on deep learning for predicting and evaluating nodes schedulable ability is proposed.Aiming at the influence of uncertainty of renewable energy generation,load and distributed energy storages on the optimal scheduling of active distribution network,a prediction model and an evaluation model of nodes schedulable ability based on deep learning are established.Firstly,in order to improve prediction accuracy of renewable energy generation and load,a node prediction model based on deep belief network is built,and is verified to have better prediction effect by comparing with other prediction models.On this basis,evaluating indexes of the degree of distributed energy storages loss and the consumptive ability of charging distributed energy storages,and the weights of each evaluation index are determined by comprehensive weighting method,then the evaluation model of schedulable ability of nodes is formed to evaluate the schedulable ability of distributed energy storages.This part serves as the theoretical basis for optimal dispatch of active distribution network.(3)A day-ahead and intra-day time scales coordinative dispatch strategy of active distribution network considering the schedulable ability of nodes is proposed.In order to solve the unreasonable problem of day-ahead dispatching plan of active distribution network and make the optimal dispatching result more in line with the actual operation of the system,a day-ahead and intra-day time scales coordinative dispatch method based on the schedulable ability of nodes is proposed.Firstly,a distributed energy storage model based on schedulable ability is established according to the evaluation model of the schedulable ability of nodes.Furthermore,under the scenario of peak load shifting,according to short-term prediction information of renewable energy generation and considering constraints of safe and stable operation of the system and energy storage operation,the first to establish a day-ahead dispatching model based on minimum operation cost of active distribution network.Then on the basis of day-ahead dispatching plan of active distribution network,an intra-day dispatching model based on ultra-short-term forecast information of renewable energy generation and minimum adjustment cost of active distribution network is established to timely adjust day-ahead dispatching plan of active distribution network.Thus,the day-ahead and intra-day time scales coordinative dispatch model of active distribution network is constructed.Finally,simulation examples show that distributed energy storages have better peak load shafting and the economical operation of the system is better.