Volume 12: Proceedings Applied Energy Symposium: CUE2020, Part 1, Japan/Virtual, 2020

Quantile based probabilistic wind turbine power curve model Keyi Xu, Jie Yan, Hao Zhang, Zhaoran Zhang, Shuang Han, Yongqian Liu


Accurate wind turbine power curves lay solid foundation for wind turbine performance evaluation and wind power forecasting, and then serve the planning and operating a low carbon energy system including renewables and electric vehicles. In order to improve the model accuracy, this paper presents the concept of quantile power curve and a quantile loss function based neural network algorithm, for establishing the proposed quantile power curve model. Based on the operational data of three wind turbines in a wind farm in China, a case study is carried out to validate the proposed model. The results show that the proposed quantile power curve model has good reliability.

Keywords wind turbine, quantile power curve, quantile loss function, neural network

Copyright ©
Energy Proceedings