Discrete Design Optimization of Detachable Mooring System for Floating Offshore Wind Turbine Based on Global Exploration Algorithm and Surrogate Models
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The fairlead chain stopper (FCS) is a newly developed detachable mooring system designed for installation on a floating offshore wind turbine. In this study, a discrete design optimization of the FCS was conducted using various surrogate models combined with a global exploration algorithm. The structural design of the FCS was assessed numerically through finite element analysis. Several surrogate models were used to enhance the convergence efficiency of the optimization procedure. The constraints for the optimization problem were defined based on strength performance evaluations. The thickness dimensions of the key components were treated as discrete or continuous design variables, enabling a comparative analysis of their influence on achieving the minimum weight, which was set as the objective function. A discrete-variable-capable global exploration algorithm was adopted to search for the optimal solution. The optimal solution for the FCS design could be determined with high accuracy using an artificial neural network-based surrogate model. A comparison of the continuous and discrete design optimum results showed relatively minor variations in the design variables across surrogate models. This study confirmed that the surrogate model accuracy plays a more critical role in determining the optimal results than the type of design variable used.
Abstract
The fairlead chain stopper (FCS) is a newly developed detachable mooring system designed for installation on a floating offshore wind turbine. In this study, a discrete design optimization of the FCS was conducted using various surrogate models combined with a global exploration algorithm. The structural design of the FCS was assessed numerically through finite element analysis. Several surrogate models were used to enhance the convergence efficiency of the optimization procedure. The constraints for the optimization problem were defined based on strength performance evaluations. The thickness dimensions of the key components were treated as discrete or continuous design variables, enabling a comparative analysis of their influence on achieving the minimum weight, which was set as the objective function. A discrete-variable-capable global exploration algorithm was adopted to search for the optimal solution. The optimal solution for the FCS design could be determined with high accuracy using an artificial neural network-based surrogate model. A comparison of the continuous and discrete design optimum results showed relatively minor variations in the design variables across surrogate models. This study confirmed that the surrogate model accuracy plays a more critical role in determining the optimal results than the type of design variable used.