DIGITAL TWIN-DRIVEN OPTIMAL DESIGN OF CONNECTOR STIFFNESS FOR FLOATING MULTI-MODULAR STRUCTURES - AN APPLICATION FOR FLOATING SOLAR ENERGY HARVEST

This paper proposes an optimization framework for connector stiffness design for multi-modular floating structures, designed for offshore floating solar in open seas. The framework integrates an optimization scheme with a digital twin. An early-phase digital twin was developed to assess the structure's response to waves as a basis for evaluating a connector stiffness parameter that mitigates environmental effects. An objective function considers module motions and internal connector loads, offering flexibility in tuning based on priorities. Bayesian optimization was adopted to find the global optimal connector stiffness, utilizing Gaussian Process regression. This method is suitable for optimizing objective functions that are expensive to evaluate, e.g. resource-intensive simulations of the response of the large multi-modular structure. The optimization process was validated and verified through model-scaled experiments and simulations in regular waves with different wave periods. In conclusion, the proposed framework demonstrates the capability to find optimal stiffnesses for different sea states, mitigating environmental effects on multi-modular structures. Future work will concentrate on improving the digital twin fidelity, reducing simulation time and developing an optimal control algorithm to enhance overall multi-modular structure performance in real-time.