|
Vol.14, No.1, February 2025. ISSN: 2217-8309 eISSN: 2217-8333
TEM Journal
TECHNOLOGY, EDUCATION, MANAGEMENT, INFORMATICS Association for Information Communication Technology Education and Science |
Modeling of Wind Speed Distribution in Urban Environment for the Application of Wind Energy Potential Estimation: Case Study
Oleksii Mykhailenko, Nadezhda Karabut, Volodymyr Doskoch, Olena Burtseva, Vitaliy Kuznetsov, Sergij Tsvirkun, Hanna Kolomits
© 2025 Oleksii Mykhailenko, published by UIKTEN. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. (CC BY-NC-ND 4.0)
Citation Information: TEM Journal. Volume 14, Issue 1, Pages 107-116, ISSN 2217-8309, DOI: 10.18421/TEM141-10, February 2025.
Received: 17 July 2024. Revised: 30 December 2024.
Abstract:
When designing wind farms, the first stage is always an assessment of the target area wind energy potential. It is necessary to have a mathematical description of the wind speeds occurrence probability at the wind turbines potential location to do this. An analysis of relevant studies shows that the most effective approach to obtaining such dependencies is when the wind speed is taken as a random variable. In this case, wind speed distribution in the target area can be modeled using continuous probability distributions. This article is devoted to determining the typical probability distribution models for representing wind conditions in certain areas of the Dnipropetrovsk oblast (Ukraine), which can be used to estimate expected level of power generation by wind power plants. To obtain the data, a series of wind speed measurements were taken at three locations throughout the year. After that, frequency wind speed distributions with ranges of 0.2, 0.5, and 0.8 m/s were created from the obtained dataset and then approximated by continuous probability distributions. Frequency distributions were modeled by Weibull, Rayleigh, Nakagami, gamma, normal, log-normal, generalized extreme value, Birnbaum-Saunders, Wald and Rice continuous distributions. To determine the parameters of each type probability distribution, which is the most relevant to the frequency distribution, the maximum likelihood estimation method was used. To assess the accuracy of the models, the Pearson test, coefficient of determination and normalized root mean square deviation are used. The probability distributions quality is also evaluated graphically using Q-Q plots. The best fit to wind speed frequency distributions demonstrated by the Weibull probability distributions. A slightly lower accuracy was provided by the normal, Rice and Nakagami distributions than Weibull distribution. But in some cases, these distributions have even smaller error than the last one. Therefore, after detailed analysis and validation, they can also be used. The Rayleigh distribution had the worst accuracy, the Pearson test for it rejected the null hypothesis that the probability distributions correspond to the frequency distributions at all three locations. Additionally, the effect of the frequency distribution wind speed grouping range on the quality of maximum likelihood estimation of continuous distribution parameters was analyzed. It showed that the approximation accuracy decreases with increasing range.
Keywords – Power system, wind energy, wind speed, modeling, probability distributions. |
|
----------------------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------------------- |