Wind energy is of main importance in the transition away from fossil fuels.
Having a yearly market growth of 15-20%, it is however necessary to face new challenges on a market that is more and more competitive. In this new and exciting context, we have been working in close collaboration with a leading energy provider in Nord Europe (Vattenfall) to explore the optimization challenges in the field. In this presentation we will give an overview of the problems we have been working in the last three years.
In particular, we used Mathematical Optimization techniques to solve large-size optimization problems arising in practical applications. More specifically, we worked on the optimization of offshore wind parks, both for the optimal allocation of wind turbines (Wind Farm Layout), as in the connection of turbines with cables (Offshore Cable Routing). Tests on real-world instances show the huge potential of using Mathematical Optimization in this (new) field of application.
Finally, a Machine Learning application in the wind sector is considered. More generally, we investigate if a machine, trained on a large number of optimized solution for a specific problem, could accurately predict the value of optimized solutions for new instances. This research question, that could potentially be interesting for many different applications, was investigated for the Offshore Wind Farm Layout problem.