Direct Parameter Optimization of Multiple DVAs for Wind Turbine Seismic Control: Comparative Performance Assessment Under Parametric Uncertainty

Wind turbine towers in seismic regions require vibration mitigation strategies ensuring structural integrity and operational safety. Multiple dynamic vibration absorbers (DVAs) offer a passive control solution, but their design traditionally relies on analytical formulations derived for simplified systems. This study systematically compares analytical versus metaheuristic approaches for DVA parameter design, evaluating nominal performance and sensitivity to parametric uncertainties. Building on previous research where DVA placement and tuning frequencies were optimized using Den Hartog analytical formulas for parameter calculation, this work uses a metaheuristic algorithm to directly optimize the mass, stiffness, and damping coefficients of five DVAs installed at previously determined optimal locations. A finite element model incorporating tower, nacelle, and rotor blades subjected to synthetic seismic excitations is employed for dynamic analysis. Two design configurations are systematically evaluated: analytical tuning based on classical formulations versus direct numerical optimization of all physical parameters. Performance is assessed under nominal conditions for acceleration and displacement at the nacelle. Subsequently, an uncertainty analysis using Latin Hypercube Sampling with three hundred scenarios evaluates both configurations under structural property variations, DVA manufacturing tolerances, and seismic variability. Statistical metrics including mean response, coefficient of variation, extreme percentiles, and degradation factors are computed to quantify the consistency and reliability of each design strategy. This comparative framework enables identification of potential trade-offs between nominal optimality and parametric sensitivity. The findings will provide quantitative guidance for practitioners selecting DVA design methodologies based on confidence in system characterization and acceptable performance variability under uncertain conditions.