Reanalysis of Jacket Support Structure for Computer-Aided Optimization of Offshore Wind Turbines with a Genetic Algorithm

Sebastian Schafhirt, D. Zwick, Michael Muskulus

The Twenty-fourth International Ocean and Polar Engineering Conference · 2014 · 15 citations · 14 references

Concepts

Abstract

The optimization of jacket support structures for offshore wind turbines is a nontrivial task. Due to nonlinear and time history-dependent effects, the analysis is simulation-based. Structural optimization using time-domain simulations is computationally demanding and difficult and typically requires a gradient-free approach. A genetic algorithm can be used for this purpose, but is limited by the computational resources available. This paper presents an approach to modifying the standard genetic algorithm for the automatic optimization of offshore wind turbines with jacket support structures under fatigue constraints. It is shown that performing a reanalysis of the jacket structures, by using performance data from earlier analyses in parallel with the simulation-based analysis process, requires a smaller number of iterations to obtain improved designs. Thus, the use of reanalysis within the genetic algorithm speeds up the algorithm significantly.

References

14