Research on Short-term Wind Power Forecasting Based on Hybrid Deep Learning
Aiming at the problem of low accuracy of existing wind power forecasting,an ultra short term wind power forecas-ting model based on IEMD and hybrid deep learning model was proposed.Firstly,the original wind power data was decomposed based on IEMD,so as to decompose the power high-frequency,intermediate frequency,low-frequency and trend characteris-tics.Secondly,the power mid frequency,low frequency and trend characteristics were predicted based on the least squares sup-port vector machine,and the wind power high-frequency characteristics were predicted based on LSTM network.Finally,the fi-nal prediction result was obtained according to the feature superposition rule.In the experiment stage,the wind power data set re-leased by a Chinese power company was used for the experiment.MAPE,MAE,RMSE and other indicators of the proposed mod-el were better.The experimental results verify the feasibility and effectiveness of the proposed model.The model provides a cer-tain reference for the intelligent service of hybrid smart grid and the application and development of new energy dispatching plan-ning.
万方数据
维普
