Design Optimization of a Tension Leg Platform for Floating Offshore Wind Turbine Based on NSGA-II Algorithm With Surrogate Models

Abstract

Abstract. This study performs the optimization of a tension leg platform for a floating offshore wind turbine (FOWT) to minimize pitch motion and tendon tension responses using time-domain numerical simulations. To achieve this, the nondominated sorting genetic algorithm II (NSGA-II) was employed in combination with surrogate models. The target platform features a central circular column and three offset square columns linked by pontoons. Initially, input design variables were selected through sensitivity tests of mooring specifications and hull shape parameters. Subsequently, a large set of sample points was generated considering the constraints, followed by individual evaluations of their hydrodynamic performance. For hydrodynamic analysis, frequency-domain and time-domain simulations were conducted using wadam and openfast, respectively. Surrogate models were constructed using the response surface method. Finally, the NSGA-II algorithm was applied to derive Pareto-optimal solutions for the multi-objective function using the surrogate models. The performances of the identified optimal solutions were verified through direct simulations. The optimized designs exhibited significant improvements in pitch motion and tendon tension responses compared to the initial design. Spectral analysis revealed that the optimal solutions mainly reduced the wave-frequency components of pitch motion and tendon tension responses. The discussion focuses on the design parameters that are most critical for improving the platform’s performance.