Scheduling Method Of Energy Storage Station With Large Scale Distributed Photovoltaic Connected Power System

With the increasing tension of conventional energy supply and the continuous rise of environmental protection calls,and relying on government financial subsidies and preferential grid-connected policies,distributed photovoltaic power generation has shown strong development potential with its clean and flexible advantages,but this has led to a significant increase in the volatility,randomness and uncertainty of regional grid load after PV grid-connected,which not only causes the difficulty of fitting the random law of the load to rise sharply,and has a large impact on the formulation and implementation of the dispatch plan,but also brings a series of new problems to the peak shaving of the power grid.At present,energy storage has high charging and discharging efficiency,short response time and flexible configuration,which can effectively dynamically transfer power and energy from time to time.It provides an effective way to solve the problem of new energy consumption and peak shaving caused by distributed photovoltaic connected power system.Based on this,in order to reasonably describe the random law of regional grid load change after photovoltaic grid-connected,a typical load scenario selection method based on improved K-means algorithm is proposed to analyze the influence of large-scale distributed photovoltaic grid-connected system on system load and peak-shaving capability.And considering the ability and advantages of energy storage power stations participating in regional power grid peak shaving,this paper studies the scheduling method of energy storage power station as an independent power supply to assist the conventional unit peaking.It provides a new way to solve the problem of peak shaving under the condition of high permeable distributed photovoltaic grid-connected.The specific research content of the paper is as follows:(1)The ability of energy storage power stations to participate in peak shaving of the system under the condition of large-scale distributed photovoltaic grid-connected is studied.This paper analyzes the factors affecting the output power of distributed photovoltaic power generation system.Then the characteristics of distributed photovoltaic output and the influence of large-scale distributed photovoltaic grid-connected on grid operation are discussed.And the basic principle of using the energy storage power station to improve the peak shaving of the power grid is studied.And for the peaking demand,it summarizes the charging and discharging capacity of various energy storage power stations.These provide theoretical support for subsequent research.(2)In order to reflect the change of load after distributed photovoltaic grid-connected,this paper proposes a method based on improved K-means algorithm for typical load scene selection.In this method,the initial cluster center number is determined by the sum of square and increment of intra-class deviation,and then the historical data of photovoltaic and load are clustered by K-means algorithm to form a typical scene set,which can reflect the characteristics of historical data.The sample data closest to the clustering center is selected as the typical daily load scenario in the typical scene.Finally,the effectiveness of the method is verified by simulation.(3)Aiming at the problem of system peaking caused by large-scale distributed photovoltaic integration,this paper proposes a scheduling method for peak shaving of conventional units using energy storage power stations.Firstly,the influence of photovoltaic grid-connected system on system load and peak-shaving capability is analyzed from the selected typical load scenario.Then,the energy storage station is used as an independent source to access the system,and this paper proposes a scheduling method for peak shaving of conventional units with energy storage power stations.Finally,the effectiveness of the method is verified by simulation.