Article,
A new approach for synthetically generating wind speeds: A comparison with the Markov chains method
Energy, (January 2013)
DOI: 10.1016/j.energy.2012.10.032
Abstract
The steady growth in the utilisation of wind power for electricity generation has led to increased interest in methods for synthetically generating wind speeds that are able to more accurately determine the site potential. The aim of this work is to develop a parametric model for generating hourly wind speed, that uses readily available aggregate statistical data, such as mean and maximum monthly or yearly wind speeds. The model has been validated using wind speed measurements collected during an experimental campaign at a site in Italy. To assess performance differences with respect to an already established methodology, the series of wind speeds generated with the proposed approach has been compared with that obtained using the Markov chains method. Various comparison criteria have been considered, including the overall statistical parameters, distribution function, autocorrelation, and power spectral density. ⺠A methodology for synthetic generation of hourly wind speed data is proposed. ⺠The proposed approach is compared with a first order Markov chain method. ⺠Model is validated using wind speed time series measurements. ⺠Comparison criteria include the distribution function and power spectral density. ⺠The proposed approach is able to preserve the daily trend of measured data.
