Natural Frequency Degradation Prediction for Offshore Wind Turbine Structures

The scale of offshore wind turbines (OWTs) has increased in order to enhance their energy generation. However, strong aero/hydrodynamic loads can degrade the dynamic characteristics OWTs because they are installed on soft seabeds. This degradation shorten structural life system; repetitive lead seabed softening, reducing natural frequency structure close excitation frequency. Most previous studies trained prediction algorithms with actual sensor signals. there no data response over lifespan (approximately 20 years). In address this issue, study proposes a new platform combining OWT model and neural network-based model. Specifically, virtual was generated using three-dimensional finite element Then, LSTM predict as input. results show that developed accurately frequencies next several years past present accelerations strains. practice, could be used future structure, which measured sensors (accelerometers strain gauges).