Comparative Assessment of Wind Speed Predictive Capability of First-and Second-order Markov Chain at Different Time Horizons for Wind Power Application

Abstract

In this article, comparison between the first-and the second-order Markov chain in the prediction of wind speed of different time horizons (i.e., very-short-term, short-term, medium-term and long-term) are performed using the wind speed data of Alexander Bay, South Africa. The prediction capability of the first-and second-order Markov chains are tested by comparing the predicted values to the measured value using standard statistical tests. Furthermore, the values of the predicted wind speeds are used to forecast wind power for each of the time horizons using ENERCON-E53 wind turbine power curve. Some of the key results reveal that both first-and second-order Markov chains present good forecast of wind speeds on very-short-time horizon and as prediction times increase, they lose their precision. Hence, they work best with very short prediction time. The result also shows that second-order Markov chains, which are based on the second-order transition matrices, have better performance compared to first-order Markov chains.