ML-Optimization Steel Structure: Offshore Wind Turbine
Please login to view abstract download link
Offshore wind energy leverages the high intensity and consistency of oceanic winds, making it essential for global renewable energy transition. As energy demand rises, the deployment of larger offshore wind turbines becomes critical for maximizing power generation and reducing average electricity costs over a project's lifespan. However, scaling up wind turbines presents significant engineering challenges, especially in designing support structures like towers. These towers must withstand increased loads while maintaining structural integrity, cost-effectiveness, and transportability. Traditional stress evaluation methods in structural engineering often involve computationally intensive techniques and are limited by their reliance on differential equations and discretization. Deep learning has emerged as a faster and more adaptable alternative. This research introduces an ML-based model aimed at optimizing the design of offshore wind turbine towers by generating new designs and predicting their physical properties. It utilizes a large dataset of floating wind turbines, integrating topology optimization and finite element analysis (FEA) to achieve designs optimized for deflection, stress distribution, modal analysis, buckling, and fatigue resistance. While focused on offshore wind turbines, this approach can be extended to other 3D steel structures.
