Adapting a deep convolutional RNN model with imbalanced regression loss for improved spatio-temporal forecasting of extreme wind speed events in the short to medium range

[摘要] The number of wind farms and amount of wind power production in Europe, both on- and offshore, have increased rapidly in the past years. To ensure gridstability and on-time (re)scheduling of maintenance tasks and to mitigate fees in energy trading, accurate predictions of wind speed and wind power areneeded. Particularly, accurate predictions of extreme wind speed events are of high importance to wind farm operators as timely knowledge of thesecan both prevent damages and offer economic preparedness. This work explores the possibility of adapting a deep convolutional recurrent neuralnetwork (RNN)-based regression model to the spatio-temporal prediction of extreme wind speed events in the short to medium range(12  h  lead time in 1  h  intervals) through the manipulation of the loss function. To this end, a multi-layered convolutional longshort-term memory (ConvLSTM) network is adapted with a variety of imbalanced regression loss functions that have been proposed in the literature:inversely weighted, linearly weighted and squared error-relevance area (SERA) loss. Forecast performance is investigated for various intensitythresholds of extreme events, and a comparison is made with the commonly used mean squared error (MSE) and mean absolute error (MAE) loss. Theresults indicate the inverse weighting method to most effectively shift the forecast distribution towards the extreme tail, thereby increasing thenumber of forecasted events in the extreme ranges, considerably boosting the hit rate and reducing the root-mean-squared error (RMSE) in thoseranges. The results also show, however, that such improvements are invariably accompanied by a pay-off in terms of increased overcasting and falsealarm ratio, which increase both with lead time and intensity threshold. The inverse weighting method most effectively balances this trade-off, withthe weighted MAE loss scoring slightly better than the weighted MSE loss. It is concluded that the inversely weighted loss provides an effective wayto adapt deep learning to the task of imbalanced spatio-temporal regression and its application to the forecasting of extreme wind speed events inthe short to medium range.

[学科分类] 土木及结构工程学