Short-term offshore wind power forecasting - A hybrid model based on Discrete Wavelet Transform (DWT), Seasonal Autoregressive Integrated Moving Average (SARIMA), and deep-learning-based Long Short-Term Memory (LSTM)

Short-term time series wind power predictions are extremely essential for accurate and efficient offshore energy evaluation and, in turn, benefit large farm operation maintenance (O&M). However, it is still a challenging task due to the intermittent nature of wind, which significantly increases difficulties forecasting. In this paper, novel hybrid model, using unique strengths Discrete Wavelet Transform (DWT), Seasonal Autoregressive Integrated Moving Average (SARIMA), Deep-learning-based Long Short-Term Memory (LSTM), was proposed handle different components an turbine Scotland, where neither approximation nor detail considered as purely nonlinear or linear. Besides, integrated pre-processing method, incorporating Isolation Forest (IF), resampling, interpolation applied raw Supervisory Control Data Acquisition (SCADA) datasets. The DWT-SARIMA-LSTM model provided highest accuracy among all observed tests, indicating could efficiently capture complex times patterns from power.