Wind power forecasting based on outlier smooth transition autoregressive GARCH model

摘要: The impacts of outlying shocks on wind power time series are explored by considering the outlier effect in volatility series. A novel short term forecasting method based smooth transition autoregressive (OSTAR) structure is advanced, then, combined with generalized conditional heteroskedasticity (GARCH) model, OSTAR-GARCH model proposed for forecasting. further to be fat-tail distribution. Consequently, mechanisms regimes against different magnitude investigated owing parameters models. Furthermore, depicted news impact curve (NIC) and a regime switching index (RSI). Case studies practical data validate feasibility method. From forecast performance comparison models, distribution proves promising