Research For Wind Power Ramp Events Prediction And Control Strategy Based On Residual Correction

Wind power ramp events refers to the wind power fluctuate violently in a short time.Wind power grade produce serious harm to the stable operation of power system in the process of grid connection.Under extreme weather conditions,the wind power ramp events will destroy the original power balance in the power grid,thus seriously affecting the normal power supply at the load end.Therefore,it is very important to predict the ramp events accurately and put forward an effective control method.Aiming at the problems of low accuracy of wind power ramp prediction leading to poor stability of the power system,this paper puts forward the prediction method and control strategy of wind power ramp.Specific research contents are as follows:(1)The Spinning Door Transformation is used to identify wind power ramp events,on the basis of which the improvement of the Spinning Door Transformation is carried out.And the feasibility and effectiveness of the improved Spinning Door Transformation to identify and extract wind power ramp events is verified by the simulation of an example.The preparation is made for the wind powerramp events prediction in chapter 3.(2)This paper puts forward a kind of based on markov switching autoregressive model(MSAR)correction of wind power prediction residual improved model.A kNN-SVR model based on k nearest neighbor is proposed to fuse three independent short-term numerical weather forecasts and reduce the uncertainty of numerical weather forecasts.Using the back propagation neural network(BPNN)to forecast the wind power,and verifying the effectiveness of the method by an example.Based on this,using markov switching autoregressive model(MSAR)of the originalmodel to predict residual correction.The improved Spinning Door Transformation is used to extract line segments.And the accuracy and effectiveness of the model are verified by example analysis and comparing with the traditional prediction model.(3)The wind power ramp rate is forecasted in this paper.Based on the establishment of multi-time series wind farm ramp rate prediction model,the improved adaboost algorithm is proposed to educe the dimension of impact factors and determine the input impact factors and their weights.Based on the improved adaboost algorithm model,the SVM algorithm is adopted to predict wind power ramp events rate.And the accuracy and effectiveness of the model are verified by example analysis and comparing with the traditional prediction model.(4)This paper studies the control strategy of wind power ramp events based on mixed energy storage.On the basis of mixed energy storage,an advanced control strategy based on MPC(model predictive control)is proposed.Forming advanced control scheme of accumulator-hydrogen hybrid energy storage system based on the optimization goal of the minimum grid-connected power fluctuation,the smoothest energy storage output and the minimum energy loss.The wind power ramp events control is realized according to the ramp prediction,and the economy of wind farm is improved.The effectiveness and feasibility of the control strategy are verified by engineering examples.