Analyzing real-world engineering problems such as wake modeling of wind/ocean current turbines are known to be complex and challenging. The multivariable nature of these problems requires either the implementation of computational analyses under certain simplifying assumptions or conducting experiments for a limited number of scenarios. Hence, there is always several fundamental features missed in understanding the key players in determining the complex turbulent velocity fields within the wake of turbines. It becomes more critical when studying the optimization of wind/ocean renewable farms with more than one turbine to determine the true power density or cost of energy.
Machine learning (ML) algorithms suggest promising complementary solutions to the existing physics-based (e.g. wind farm wake modeling) techniques. Implementation of conventional ML algorithms that require long-term historical data is either not feasible in many real-case applications or very expensive and time-consuming. Moreover, there are often infinite features in dataset with complex relation between them. It makes the tasks of feature selection and model tuning more challenging. In this work, a cross-domain study of physics and ML models is performed to show the need of integration of these domains. The key achievement of this work is two-fold: first, suggesting a group of emerging generative models (e.g. Generative Adversarial Networks) in the wake modeling domain; second, reducing the computational cost by demanding either smaller or no simulation dataset.