Research On Probability Interval Prediction Of Ultra Short Term Wind Power Based On Granular Computing

In the context of global joint efforts to promote sustainable development and green development,the role of wind power technology in the field of electric energy is becoming more and more important.However,wind power has long been criticized for its adverse effects on the stability of power grid due to its randomness,volatility and uncertainty.Therefore,the prediction of wind power has great impact on the healthy development of the whole wind power industry.Based on the measured data of wind farms in the northeast of China,the related research is carried out from the abnormal data of wind power,the characteristics of the time series fluctuation of wind power,the ultra short term prediction of wind power and the prediction of the ultra short term probability interval of wind power.First of all,in view of the complicated and complicated algorithm of wind power anomaly recognition,and the poor recognition effect,and the difficulty in identifying the complex and changeable wind power data,a different wind speed range under the different wind direction is proposed according to the characteristics of the abnormal data sources in the measured wind power data.Identification of internal abnormal data.The algorithm identifies the corresponding outliers based on the optimal variance in different wind speeds under different wind directions.Secondly,through the analysis of the measured power data samples from the wind farm,an analysis method of wind power time series fluctuation characteristics based on the local extreme difference rate is proposed,and some continuous output states of wind power are extracted to describe the continuous wave characteristics of wind power.In order to measure the amplitude and phase angle of the local extreme difference rate as the model input,a grey multi-objective decision model is established,and then the fluctuation coefficient is defined to quantify the fluctuation of the wind power in a period of time.Finally,the evaluation index of the prediction and prediction of the current wind electric field using the wave coefficient is given.The way.benefitThirdly,an ultra short term prediction model of wind power based on granular computing is proposed.The particle calculation can assemble objects with similar attributes and obtain different information particles according to the needs of each particle level,and then the solution can be transformed to the different granularity level;from this,the original data can be obtained.The basic characteristics of the system are fully preserved,and the transparency of the system is also improved.This lays the foundation for the realization of the ultra short term prediction of high-precision wind power.At the end of this paper,a method of wind power probability interval prediction based on Copula theory is proposed.Based on the correlation between the actual value of wind power and the predicted value,the correlation between the prediction value of wind power and the actual value is analyzed by using the related theory of Copula function,and the wind power is calculated under the condition of a certain value.The conditional probability distribution of the actual value of the power is transferred to the conditional probability analysis of the error,and then the estimation of the distribution of the error is converted to the uncertainty estimation of the wind power prediction.