Time Series Analysis and Forecasting of Wind Speed Data

Abstract

This paper discusses the problem of predicting wind speed using the statistical model based on autoregressive integrated moving average (ARIMA). Historical wind speed data, representing the Chester region of Nova Scotia, Canada, from 2012 to 2017, was used to operate this model. The form structure is defined by the rows p, d, q, and the length of the data period retrospectively. The structure parameters, autoregressive and moving average, were determined by the partial auto-correlation function and auto-correlation function, respectively. The model forecasting accuracy is based on the root mean square error, the mean absolute percentage error and the mean absolute error.