Host: Japan Association for Wind Engineering
Pages 85-86
A gust forecasting model by based on ARX model which uses numerical weather prediction and on-site measurement as inputs, was proposed and the model parameters were estimated by using non-parametric regression with forgetting factor. By using dynamically adaptive model, the prediction accuracy was improved compared to the conventional static MOS model. It was also shown that the prediction accuracy of maximum gust improves by utilizing the model ensemble method which uses two different numerical weather predictions. The predictability of the gust event with the maximum wind speed larger than 15ms, was evaluated by using ROC curve and AUC and it was shown that proposed method improves predictability of gust wind event.