An active control scheme based on compensation of perturbation force signals is proposed to balance rotating machinery with unknown system parameters and variable operation speed. The algebraic parameter identification methodology is applied to estimate the mass, stiffness, damping, rotor eccentricity, and on‐line reconstruction of the unknown centrifugal forces induced by rotor unbalance. In addition, an asymptotic estimation scheme of perturbation signals, which requires position measurements only, is proposed as another alternative to estimate such unbalance forces. Our design methodology considers that a disturbance signal can be locally approximated by a family of Taylor polynomials. Some simulation results are included to show the dynamic and robust performance of the active balancing scheme and perturbation force estimation methods proposed for a rotor‐magnetic bearing system.