Standardized Software for Wind Load Forecast Error Analyses and Predictions Based on Wavelet-ARIMA Models -- Applications at Multiple Geographically Distributed Wind Farms

This study introduces a standardized forecast error analysis and prediction tool that can be implemented into a software package that predicts the uncertainty range for generation resources involved in the power grid balancing service. The tool is reusable and transferrable to power generation systems of different variability. Given the multi-scale variability and uncertainty of wind generation and forecast errors, a natural choice is to use time -- frequency representation (TFR) as a view of the corresponding time series represented over both time and frequency. Here we use wavelet transform (WT) to expand the signal in terms of wavelet functions that are localized in both time and frequency. Each WT component is more stationary and has a consistent auto-correlation pattern. We combined wavelet analyses with time-series forecast approaches such as the autoregressive integrated moving average (ARIMA) model and tested the approach at five different wind farms with similar or different field and weather conditions. The prediction capability is satisfactory -- the day-ahead prediction of errors matches the original error values very well, including the patterns. The observations are well located within the predictive intervals. Integrating our wavelet-ARIMA ("stochastic") model with the weather forecast model ("deterministic") will improve our ability significantly to predict wind power generation and reduce predictive uncertainty.