Evaluation of NeuroSim-FLOAT: A Multi-Agent AI Framework for Fatigue-Aware Design Optimization of Floating Offshore Wind Turbine Towers
Please login to view abstract download link
Artificial intelligence is reshaping how complex engineering systems are designed and simulated. Recent advances show how large language models and multi-agent coordination can automate tasks that previously required extensive expert effort. In aerospace and automotive design, multi-agent AI systems are already being explored for coordinating aerodynamics, structures, and controls. Offshore wind, however, has yet to embrace these approaches, despite being a highly multidisciplinary engineering domain. Scaling toward 20–25 MW turbines requires tight integration of aerodynamics, hydrodynamics, structures, controls, and operations. High-fidelity methods such as CFD, FEA, and coupled co-simulation provide accurate predictions but remain computationally expensive. NeuroSim-FLOAT is introduced as a multi-agent AI framework specifically tailored to floating offshore wind turbine structural component design, simulation, and optimization. The framework builds upon the previously developed FLOAT methodology, which demonstrated fatigue-aware redesign of the IEA 22 MW floating wind turbine tower through a largely manual and sequential workflow. NeuroSim-FLOAT generalizes this approach by replacing manual orchestration with autonomous AI agents, each responsible for numerical simulation, frequency-domain analysis, fatigue evaluation, and design optimization, coordinated through shared objectives and negotiation. A proof-of-concept study demonstrates that NeuroSim-FLOAT can automatically translate high-level design intents into solver-ready simulations, execute physics-based analyses, and converge to physically consistent design solutions without manual intervention inside the design loop once initialized. Compared to the original FLOAT workflow, the proposed framework employs a coordinated system of four specialized AI agents, to automate workflow coordination while preserving multidisciplinary fidelity. By integrating physics-based solvers with cooperative AI agents, NeuroSim-FLOAT enables scalable and transparent design workflows, providing a practical pathway toward faster design cycles and reduced cost of energy for next-generation floating offshore wind turbines.
