Quantum-Inspired Adaptive Control and Multi-Objective Optimization for Floating Offshore Wind Turbines Under Time-Varying Aero-Hydrodynamic Uncertainties

Abstract

Offshore wind farms are moving into deeper water to capture stronger winds. Floating Offshore Wind Turbines (FOWTs) encounter severe operational challenges. Wind and water loads shift constantly, and loads have different frequency bands, resulting in changing operating conditions. Controlling these turbines becomes nearly impossible. Adaptive control has improved, but one problem stays unsolved; no system has been built that guarantees stability mathematically while also using AI to handle unexpected disturbances in real time on FOWTs. On top of that, automatically tuning multiple control settings under unpredictable wind and wave conditions was barely explored. This paper builds a three-layer system inspired by quantum computing. Layer one uses feedback linearization and sliding-mode control for steady baseline tracking. Layer two runs quantum-inspired attention mechanisms to detect and counteract unmeasured wind and water disturbances as occurrence. Layer three uses a Quantum-Inspired Evolutionary Algorithm to optimize all the control settings offline without gradients. We tested scenario on the NREL 5 MW turbine using real wind turbulence and wave spectra from IEC 61400-1 Class B standards and JONSWAP measurements. For each of 30 runs, with different random seeds, we achieved an average rotor-speed error of $0.0267\pm 0.0043$ rad/s for our Quantum-Hybrid controller. This is 6.6 times more efficient than the simple PID controller, at 85%. PID scored $0.1767\pm 0.0670$ rad/s. Wilcoxon test shows it is statistically with p $= 1.86\times 10 ^{-9}$ . The error reduced to 0.0141 rad/s during a 600 s test run. The platform surge decreased by 7.8% and the control effort decreased by 5.6%. After 22 generations, our evolutionary algorithm determined the optimum settings, and after 45 generations, the fitness value was near 7.97, when optimization was complete. Quantum-inspired control runs in under 2 milliseconds per step and mathematically is proven stable using Lyapunov methods. It statistically sound across many tests, and is ready to deploy on real floating offshore wind turbines dealing with complex, shifting conditions.