Wind speed forecasting at different time scales: a non parametric approach

TL;DR

This paper introduces a nonparametric semi-Markov chain model for accurate wind speed forecasting across various time scales, outperforming simple persistence models and maintaining prediction quality over long horizons.

Contribution

The paper presents a novel indexed semi-Markov model for wind speed prediction, capturing statistical behavior and enabling reliable forecasts over different time scales.

Findings

Accurately reproduces wind speed statistical behavior

Maintains prediction quality over long horizons

Outperforms persistence models in forecasting accuracy

Abstract

The prediction of wind speed is one of the most important aspects when dealing with renewable energy. In this paper we show a new nonparametric model, based on semi-Markov chains, to predict wind speed. Particularly we use an indexed semi-Markov model, that reproduces accurately the statistical behavior of wind speed, to forecast wind speed one step ahead for different time scales and for very long time horizon maintaining the goodness of prediction. In order to check the main features of the model we show, as indicator of goodness, the root mean square error between real data and predicted ones and we compare our forecasting results with those of a persistence model.