Font Size: a A A

Short Term Load Forecasting And Scheduling Model Of Microgrid Based On Wavelet Analysis

Posted on:2020-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:S Q LaiFull Text:PDF
GTID:2392330590984232Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
The new power supply mode which combines micro-grid with distribution network is the demand of electric power development at present,and is one of the important measures to solve the difficulty of new energy utilization.Short-term load forecasting is the basis of economic dispatch of micro-grid system.Its forecasting accuracy directly affects the accuracy of economic dispatch.Therefore,it is necessary to study short-term load forecasting of micro-grid.In addition,the waste rate of new energy is very high when it is connected to the distribution network.It is necessary to study the micro-grid with different permeability of new energy for further fine modeling of the micro-grid with battery.Therefore,from the perspective of shortterm load forecasting,this paper studies the economic dispatch of micro-grid with different new energy penetration.In this paper,the power load of a certain area distribution network is taken as the research object,and the load,wind power and photovoltaic data of the area are analyzed in depth.The prediction adopts the screening mechanism of similar days.Firstly,the data is normalized,and the similar days are selected by fuzzy clustering based on the meteorological data of the past five years.According to the load data,it is divided into economic component,periodic component and random component.The economic growth component is obtained by regression analysis,and the periodic component and random component are decomposed by wavelet transform.After filtering out the irregular high-frequency component,BP neural network algorithm is used to predict.Finally,the forecast result is obtained by inverse wavelet transform.The results show that the maximum prediction error is not more than 6%.Similarly,similar measures are used to predict photovoltaic and wind power generation.After obtaining the power generation data of various loads,photovoltaic and wind power,the economic dispatch model of the microgrid is established.Firstly,the concept of virtual micro-grid is established,which links distribution network,storage battery,wind and solar power generation device and load through virtual micro-grid,eliminating the complex relationship of power flow among the various parts.Based on the detailed economic dispatch modeling of the components in the microgrid,the economic dispatch model of the microgrid with the lowest power supply cost as the objective function is constructed.Taking the model of a microgrid in a certain area as an example,the economic dispatch model includes battery dispatch,so it is changed from linear programming(LP)model to mixed integer programming(MILP)model.In this paper,the battery life is fully considered and the dispatch planning of microgrid is carried out.In addition,in order to improve the penetration rate of new energy,this paper studies the power supply cost of micro-grid under different windto-solar ratios,and gives the optimal ratio of wind-to-solar according to the use cost of storage battery.
Keywords/Search Tags:wavelet analysis, virtual microgrid, mixed integer programming, new energy penetration
PDF Full Text Request
Related items