|Description :||In the context of global climate change, the national and global energy transition aims to control and exploit several renewable energy sources, in particular offshore wind power. This industrial sector faces many scientific and technological challenges, particularly in view of the floating wind turbine technologies development. Several current issues concern the optimization of energy production. The wind turbine production in a floating wind farm depends essentially on its position, the wind and the state of the sea. The influence of the state of the sea, in particular, is poorly known, because it has not been an influential factor in the development of installed wind farms. First o all, the influence of the movement of the platform directly affects the production capacity of the wind turbine. But beyond that, it induces significant modifications to the wake, and in the context of commercial farms, where the turbines are not widely spaced, leads to interactions between turbines which are still unknown.
To deal with these recent challenges, the currently proposed strategies are mainly based on the attempt of direct numerical simulation of the problem. Nevertheless, such approaches are extremely expensive, and do not allow, beacuse of numerical cost, to reach an understanding, or a complete characterization for all the required environmental conditions (wind, sea state, thermal stratification, humidity).
To bypass this problem, from numerical simulation data as well as historical data, the environmental conditions can be modeled by stochastic models which are faster to simulate. Even if these models do not achieve the accuracy of physical models, the fac that they incorporate uncertainty into the modeling, it allows them to capture the most important characteristics of the underlying phenomena. Thus, by combining these stochastic models with the physical model, it is possible to evaluate the uncertainty relating to the key parameters (extractable energy, wake topology).|