Short-term Wind Power Portfolio Prediction for Wind Farms Based on Meta-learning
A meta-learning method using a deep convolutional neural network-based approach is proposed to address the problem that a single model is affected by the high uncertainty and volatility of wind power and the inability to match well the sequential characteristics of different wind turbines,thus resulting in low prediction accuracy.The method first automatically learns a feature representation from the original data,and then associates the learned features with a set of weights using a neural network,and assigns these weights to a set of base prediction models,resulting in a combined prediction.The experimental results show that the combined meta-learning-based method can effectively improve the prediction accuracy of wind power in the prediction evaluation on real data.
wind power predictiondeep learningmulti-step predictionmeta-deep learningcombinatorial predictor
NETL
NSTL
万方数据
维普
