Wind Farm Layout Optimization using Genetic Algorithms with a Gaussian Wake Model

In this work we have designed and implemented different Genetic Algorithms especially adapted to the Wind Farm Layout Optimization (WFLO) problem, with the goal to optimize the overall power output and electricity cable length of different Wind Farms currently in operation in Europe. The used flow dynamics framework relies on a wind turbine wake model (EPFL, 2014) that has shown a higher accuracy representing the turbine wakes, compared to the traditionally used wake models. We obtained a 16-18% reduction of the electricity cable length, when increasing the overall wind power output 0.24-0.89%. In addition, the methodology is shown to outperform the other optimization methods in literature applied to similar cases.