Complex Terrain Wind Farm Super Short Term Wind Resources Prediction Research

Wind speed is intermittent and has the character of stochastic volatility,in complex terrain,affected by the topography and elevation,the speed change will be bigger.As a result,wind power output is not stable,which brings challenges to the power grid scheduling and difficulties in wind power interconnecting.Super short term wind resource prediction in wind farm is one of the effective ways to solve this problem,it will be helpful to the active power control of the wind farm,to provide refference for reasonable scheduling plan,ensure the stable and safe operation of wind electrified wire netting.In this paper,a wind farm in Inner Mongolia is chosen as study area,the wind field history data is chosen as the research object,by researching these data the wind field resources are predicted and analyzed in short-term.Firstly,by using time sequence method,time sequence ARMA(Auto-Regressive and Moving-Average)model respectively for the forecast of wind resources in different time scales analysis,we found that either individually using time sequence method or time sequence ARMA model to forecast the result,there was always large error,and with time scale went longer,the error became larger.But compared to the time sequence method,the time sequence ARMA model prediction showed better accuracy and prediction stability.Secondly,based on time sequence,a method of super short term wind speed forecasting was given.This method which considering the random component used wavelet transform and time sequence method.By using wavelet transform to wind speed time sequence layering,which can be divided into high frequency change of wind speed time sequence and low frequency change of wind speed time sequence,the high frequency change of wind speed time sequence using time sequence ARMA model to forecast,sustained low frequency change of wind speed time sequence using time sequence method to predict.Finally the forecasting results were modified by stacking with a random array.The results showed that this method has much more accuracy and stability in the simulation predictions.Finally,a new method was given based on combining wavelet transform and time sequence method considering the random component of super short term wind speed forecasting method and kriging interpolation method which called time and space forecasting method.The new method contained both the advantages of high precision in wavelet transform and time sequence method and the advantages of fully considering theterrain in kriging interpolation method.This method can be used to predict any wind speed at any space and any time in the future within a short time,which not only realizing wind speed forecasting on the time dimension,but also realizing the wind speed prediction on the three dimension space.By using this method of wind field simulation,more study and analysis will be done for whole complex topography,which not only can better solve the problem of wind power dispatching interconnection,but also will provide reference for complex topography and wind field micro location.