Circular Statistics: An Analysis of Wind Direction Data

摘要

An interest on renewable energy had encouraged some researchers to do research on wind energy. The focus of this paper was on the fitting of wind direction data which was one of the most important aspects in renewable energy. The main objective was to find the best distribution that can displayed the wind direction data. Three circular distributions were considered and they were Von Mises, Wrapped Cauchy and generalized Von Mises distributions. Data on the average daily wind direction were collected from Solar Energy Research Institute (SERI) in Universiti Kebangsaan Malaysia. It was fitted and compared to describe the best distribution. The suitability of the distribution was judged from the value of the mean circular error. Result showed that the Von Mises distribution fit the data very well compared to the generalized Von Mises and Wrapped Cauchy distribution. Thus, the Von Mises distribution can be used to model the wind direction in this particular station.