Markov Chain Method And Its Application In Probability Prediction Of Wind Power

Due to the maturity of wind power technology,the capacity of wind power grid connection has increased greatly,and wind power grid connection has a significant impact on the stable operation of power systems.Therefore,high-precision prediction of wind power is of great significance to economic and social development.After considering the variation characteristics of wind power,the Markov chain method is introduced,and the probabilistic interval prediction analysis of wind power is based on Markov chain method,which greatly improves the anti-risk ability of single point prediction.This method can meet the needs of decision makers in decision planning and reliability assessment by quantifying the uncertainty information contained in the predicted values.The main research work of this paper is as follows.Firstly,the current status and development trend of wind power generation at home and abroad,the significance of wind power generation probability research,the application range of Markov chain method,the probabilistic prediction results of wind power and the application of Markov chain method to wind power are pointed out.The significance and engineering value of power forecasting.Secondly,in order to reduce the dependence on the existing point prediction model,the range of wind power under different confidence levels is obtained without assuming the error distribution of the predicted power.An empirical mode decomposition(EMD)is proposed,which is weighted.Wind power probability interval prediction method for Markov chain(WMC)and quantile regression(QR).In this study,EMD is applied to wind speed time series decomposition,and the decomposed wind speed component is modeled by WMC to improve the accuracy of predicted wind speed,and the QR prediction model is used to obtain the upper and lower limits of wind force at a given probability level.Next,considering that the wind power prediction is significantly correlated with its historical power,the Markov chain method is first applied to obtain the predicted power under the unsynchronized length,and then the fuzzy rough set method is used to determine the weight coefficient to calculate the predicted power,and the prediction will be obtained.The power input non-parametric kernel density(KDE)prediction model obtains the upper and lower quantiles of the prediction error probability density function at a given confidence level,and obtains the final predicted power interval in combination with the actual power.This method reduces the subjective selection of the model selection.Sexuality,better grasp the dynamic structure of sample sequences,improve prediction accuracyFinally,considering that wind speed data modeling is important for wind research,it provides valuable insights.The classical Markov chain method uses probability distributions to estimate statistical parameters.This model lacks time-varying properties and ignores the cross-dependencies between other models.Therefore,after considering the inherent dependence between power and wind speed,temperature and pressure,the hidden Markov chain(HMM)model is used to link these quantities,eliminating the necessity and correlation of direct sampling.The accuracy of the interval is finally introduced into the particle swarm optimization algorithm.The interval coverage and the prediction interval bandwidth are used as the optimization objective function.The simulation results show that the method is feasible and superior.