Adaptive feedforward cancellation strategy for unknown and time-varying disturbance with changing plant
【摘要】：This paper addresses an extended gradient-based adaptive feedforward cancellation(AFC) approach for attenuation or rejection of unknown and time varying disturbance without reference available for changing plant. In this study, an extended AFC(E-AFC) algorithm is proposed based on gradient approach. Convergence property of the E-AFC is investigated by using averaging theory and linear approximation. By introducing the division operations of magnitude and plant, improved E-AFC(IE-AFC) algorithm is derived and its performance dependence on the unknown disturbance magnitude and plant property is removed compared to the E-AFC algorithm. Numerous simulations are conducted to evaluate its superior performance compared to the conventional algorithms.