Research On Stochastic Optimal Dispatching Of Power Systems With High Proportion Of Renewable Energy

With the issue of energy shortage and environmental pollution becoming more and more serious,governments are devoted to developing renewable energy including hydropower,wind power,photovoltaic power and so on.This has transformed traditional power systems into power systems with high proportion of renewable energy step by step.But there are risks in power system dispatch operation with the integration of large-scale wind power and photovoltaic power because of wind and solar energy’s randomness and volatility.Therefore,the uncertainty of wind and photovoltaic power output is fully considered,and research on stochastic optimal dispatching of power systems with high proportion of renewable energy has been done in this paper considering reducing the system operation cost and promoting renewable energy accommodation under the premise of ensuring the reliability of power supply.Firstly,four output models of power supply including wind power,photovoltaic power,hydropower and conventional thermal power are established in this paper.The actual force of wind power and photovoltaic power is equal to the sum of the forecast force and the forecast errors,based on the point forecast data of wind power and photovoltaic power.And General Gaussian Mixture Model is used to fit the probability density distribution of the wind power and photovoltaic power forecast error.Then,the probability characteristics and the time series feature of wind power and photovoltaic power’s forecast error are analyzed.This paper has proposed a chance-constrained programming method considering correlation of random variables in view of the phenomenon that the traditional chance-constrained model of the system spinning reserve capacity ignores the correlation of wind power and photovoltaic power forecast error.The main process of the new method is that simulating several forecast error series according to the initial forecast error body and Autoregressive Moving Average Model are obtained,which are used to determine the envelope of these forecast error series.And then,confidence level series considering correlation of forecast error can be obtained.At last,a weekly optimal scheduling mode and optimal operation strategy are put forward.Also,Inflows and rolling modes which are the impacts of weekly optimal scheduling are analysed.Case studies show that the proposed weekly optimal scheduling strategy can allocate the water resources of cascade hydropower stations properly,coordinate the monthly electricity completion and promote renewable energy accommodation when the inflows are large during the water storage period.