PhD scholarship on Fast and accurate machine-learning surrogates of atmospheric flow dynamics with applications in wind energy | Kongens Lyngby, Lyngby‐Taarbæk Municipality, Denmark | Technical University of Denmark

Position:	PhD scholarship on Fast and accurate machine-learning surrogates of atmospheric flow dynamics with applications in wind energy
Institution:	Technical University of Denmark
Location:	Kongens Lyngby, Lyngby‐Taarbæk Municipality, Denmark
Duties:	To develop efficient dimension reduction methods to approximate wind farm flows across various operational and atmospheric conditions; To generate synthetic turbulence based on stochastic model or deep-learning techniques; To construct regression models capturing changes in turbulent structures for the various conditions; To compare different machine-learning techniques in terms of accuracy and computational efficiency, e.g. linear and non-linear methods; To perform detailed validation and error estimation of the models; To provide physical interpretation of the constructed models; Participate in scientific conferences and publish results in scientific journals
Requirements:	A background in data science, physics, engineering, or similar; Experience developing and using machine-learning techniques, e.g. neural networks; Ability to work with large data sets; Scientific programming experience, e.g. Python; Understanding of fluid mechanics, turbulence, boundary-layer flows and/or time series analysis is beneficial; Clear and concise communication skills in English; Positive attitude, a strong drive, critical thinking, and an eagerness to learn

Text:	PhD scholarship on Fast and accurate machine-learning surrogates of atmospheric flow dynamics with applications in wind energy To develop efficient dimension reduction methods to approximate wind farm flows across various operational and atmospheric conditions; To generate synthetic turbulence based on stochastic model or deep-learning techniques; To construct regression models capturing changes in turbulent structures for the various conditions; To compare different machine-learning techniques in terms of accuracy and computational efficiency, e.g. linear and non-linear methods; To perform detailed validation and error estimation of the models; To provide physical interpretation of the constructed models; Participate in scientific conferences and publish results in scientific journals A background in data science, physics, engineering, or similar; Experience developing and using machine-learning techniques, e.g. neural networks; Ability to work with large data sets; Scientific programming experience, e.g. Python; Understanding of fluid mechanics, turbulence, boundary-layer flows and/or time series analysis is beneficial; Clear and concise communication skills in English; Positive attitude, a strong drive, critical thinking, and an eagerness to learn