Abstract:
As a common and extensive datum to analyze wind, wind rose is one of the most important components of the meteorological elements. In this study, a model is proposed to establish the joint probability distribution of wind speed and direction using grouped data of wind rose. On the basis of the model, an algorithm is presented to generate pseudorandom numbers of wind speed and paired direction data. Afterward, the proposed model and algorithm are applied to two weather stations located in the Liaodong Gulf. With the models built for the two cases, a novel graph representing the continuous joint probability distribution of wind speed and direction is plotted, showing a strong correlation to the corresponding wind rose. Moreover, the joint probability distributions are utilized to evaluate wind energy potential successfully. In cooperation with Monte Carlo simulation, the model can approximately predict annual directional extreme wind speed under different return periods under the condition that the wind rose can represent the meteorological characters of the wind field well. The model is beneficial to design and install wind turbines.