Research On Optimization Of Wind Speed Prediction Based On Combined Model

The energy revolution is one of the strong driving force for social progress and the development of renewable energy will bring immeasurable economic and environmental benefits for all humanity.Wind energy is inexhaustible,environment-friendly and has great power generation potential.Since the early 20 th century,wind power generation has been constantly attempted.Today,it has been successful in the world and plays an important role in the field of energy.However,due to the strong volatility and intermittence,it is difficult to achieve a smooth grid connection of wind power,which is a huge obstacle to its development.Accurate wind speed prediction is the key to solve these problems above.The statistical analysis and inference of wind speed time series are the main research methods of wind speed prediction.In this paper,based on the above background,the short-term wind speed prediction of wind farm is studied.Ensemble empirical mode decomposition is a signal decomposition method in the time dimension,which can effectively reduce the complexity of the wind speed sequence and thus improve the prediction accuracy.In this paper,the application in wind speed prediction of ensemble empirical mode decomposition is analyzed.Then,a selection and comparison method based on grey relational analysis is proposed for the parameters selection of ensemble empirical mode decomposition.The multistage increasing trends are definited in this paper and the relationship between trend items and original wind speed series is analyzed through grey relational analysis.Then,according to the principle of parameters selection proposed in this paper,the results of decompositions under different initial parameters are compared and thus the relative optimal parameters can be selected.The case analysis and the related research results show that the method proposed in this paper is logical and operational,having a great practical value.The grey prediction model and the neural network method have some limitations for wind speed prediction.A combined prediction model based on ensemble empirical mode decomposition is proposed in this paper.The prediction results of wind speed components are obtained by using different prediction methods and added,through which the second order grey prediction method and neural network method are organically integrated.Case analysis shows that the combined prediction model proposed in this paper has good performance,which overcomes the shortcomings of single prediction model to a large extent,significantly improving the accuracy of wind speed prediction.In order to modify the prediction error of the combination model and further improve the accuracy,the Lorenz disturbance is added to the combination prediction model in this paper.A linear combination method is analyzed.On this basis,a new combination method using RBF neural network is proposed.The example analysis shows that the prediction accuracy of the model is improved effectively by adding Lorenz perturbation and the disturbance combination method proposed in this paper has better correction effect,wider application range,stronger operability,with great application prospect.This paper processes comparison and selection for parameters of ensemble empirical mode decomposition by using grey relational analysis;The combination wind speed prediction model is constructed by using the ensemble empirical mode decomposition and the grey prediction model is combined with the neural network model organically;The combination method of combination model and Lorenz disturbance is analyzed and a combination method through RBF neural network is proposed.