Purpose: <p>Active mass drivers (AMDs) are effective for reducing vibrations of structural systems under excitations such as wind and earthquakes. However, their control remains challenging due to uncertain dynamics, un-modeled high-frequency perturbations, unreliable predefined models, and natural disturbances. This study proposes a novel adaptive intelligent control approach that minimizes reliance on predefined models and adapts in real time to perturbations.</p> Methods: <p>The dynamics of the system are modeled online as a first-order fractional-order nonlinear system, where the fractional derivative order is adaptively updated using an Unscented Kalman Filter (UKF). Type-2 fuzzy systems (T2FSs) identify nonlinearities in the fractional-order model. Fractional Lyapunov-based adaptation laws minimize estimation errors and guarantee stability. Additionally, a supplementary parallel controller is designed to manage uncertain bounds of nonlinearities, with these bounds adaptively estimated through an adaptation law.</p> Results: <p>Numerical simulations and experimental studies validate the proposed algorithm. Results demonstrate superior vibration reduction compared to conventional controllers. The method achieves up to an 85% reduction in peak floor displacement under impulse excitation and maintains stability under seismic and harmonic disturbances where traditional methods fail.</p> Conclusions: <p>The proposed adaptive intelligent control strategy enhances vibration attenuation and structural resilience by enabling real-time adaptive modeling and robust handling of nonlinearities and uncertainties. These improvements establish the method as a more effective solution than conventional approaches for structural vibration control.</p> Supplementary material: <p>A video demonstration of the implementation is available at&#xa0;<a href="https://youtube.com/shorts/fWjeZ2uXdh8?feature=share">https://youtube.com/shorts/fWjeZ2uXdh8?feature=share</a>.</p>