Research On Distributed Energy Resource Aggregation And Optimal Operation Of Novel Power System Considering Flexibility Support

Under the background of multiple pressures such as the depletion of fossil energy,the intensification of greenhouse effect and environmental pollution,the electricity generation and control technology of renewable energy has made continuous development and progress.Plenty of centralized and distributed renewable energy sources are connected to the power system,The traditional power system has gradually developed into a novel type of power system with renewable energy as the main power source.The widespread access of renewable energy with fluctuating and intermittent characteristics brings challenges to the safe and stable operation of the power system.The novel power system with high proportion of renewable energy is suffering from flexibility scarcity problem.It is an urgent request to resolve the flexibility scarcity of novel power system.To improve the power system flexibility,research can be carried out from the power generation side,the grid side,the user side and the market mechanism.Compared with adding flexible power source or improving the flexibility of the traditional power source,it is a more economical and efficient adjustment method to improve the flexibility of user side with a wind range of distributed energy resources and adjustable load.Therefore,this paper focuses on the user-side flexibility mining technology with a wide range of distributed energy resources and adjustable load,which include the distributed energy resource and load forecast,distributed energy resource aggregation modeling method and flexible support operation.The main research contents of this paper are as follows:To improve the operation flexibility of power system,the foundation is to improve the renewable energy output forecast accuracy.Considering the forecasting problem of renewable energy power,this paper takes wind power forecast as the research object.A comprehensive similarity measurement of characteristic variables based on data mining is developed,which combines of euclidean distance and cosine angle together.Relief F algorithm is carried out to assign weight of associated feature attributes.This optimized clustering method is used to identify the historical wind speed days with similarity both in spatial distance and curve shape.In order to improve the response speed and calculation accuracy of the algorithm,the neural network with gated recurrent unit is used to predict the short-term wind power.Finally,the correlation index of system flexibility is calculated in a simulation case,it confirms the effectiveness of the proposed prediction method as a support for system flexibility adjustment capability.To achieve a balance between supply and demand for system flexibility,another important task is to improve the accuracy of user load forecast,and analysis the potential demand response accurately.Aiming at the forecasting problem of distributed user load,this paper takes multi type building load as the research object.A wide area of impact factors index(meteorological parameter,user comfort index and air quality parameter)with information mining is constructed.The best candidate input features for different types of building load are selected by a max relevancy and min redundancy criterion.The long-short term memory-recurrent neural network model is used to forecast the short-term load of different types of buildings.By calculating the flexibility correlation index of the system under the dual uncertainty resource access of source and load,it is proved that the proposed prediction method has played a positive role in the balance adjustment of supply and demand flexibility.A second-order thermodynamic physical model of building air-conditioning load considering the influence of indoor heat source parameters is further established,and the response characteristics of building air-conditioning load under the influence of indoor temperature regulation range,response period and outdoor temperature are quantitatively analyzed.Distributed resource aggregator is an effective aggregation mode of distributed resources.It can aggregate and optimize a large of internal uncertain resources and controllable resources in power system demand side.In view of the complex physical model and low decision efficiency of existing aggregators optimizing demand response resources,the effective input variable-output decision data structure relationship is selected based on the individual response characteristics analysis.The demand response characteristic encapsulation model of aggregator is obtained by using the deep learning method with Inception structure.Through case simulation,it is verified that the proposed method can schedule demand response resources effectively and rapidly.It is shown that the method can improve the flexibility of aggregator to aggregate distributed resources,provide technical support for their participation in the flexible operation of the power grid.Distributed resource aggregator in the user side can provide strong support for the power system flexible operation ability improvement.An effective electricity market transaction mechanism is designed,aiming to fully activate the potential ability of flexibility resources.The lower-level distributed aggregator participates in the distribution network transaction as an encapsulated and modularized benefit individual,which avoids the tediousness of repeatedly iterating,improves the efficiency of the system optimization calculation.The upper-level distribution network operator builds a multi-objective dayahead optimal operation model that comprehensively considers the correlation indicators of economy and flexibility.The objects of regulation include distributed power generation,energy storage and distributed resource aggregators.NSGA-Ⅱmethod is used to find the Pareto optimal boundary domain set,and the optimal day-ahead result is determined based on the TOPSIS method.The case simulation proves the effectiveness of the method proposed in this paper to improve the economy and flexibility of power system operation.