Adaptive tracking control of dual position loops drive using active-disturbance- rejection control
Baozhu Xue,Chengyong Zhang,Longfei Zhang,Miannuo Chen,Yaolong Chen
2019-01-01
Abstract:To eliminate the influence of the internal and external disturbance, a novel active-disturbance-rejection controller (ADRC) for dual position loops feed-drive system has been proposed in this paper. This controller is based on the disturbance-rejection tracking control (DRTC) and proportion-integration-differentiation (PID) control. The controller adopts a dual position loops feedback control scheme which is different from the traditional feed system with position loop and speed loop. The motor rotation angle position is used in the inner loop as the feedback signal to reduce the influence of differential error, noise amplification and filtering delay. The closed-loop control of the whole system is realized in the outer loop through the feedback of the load position. Both of the position loops have the position controller and the state observer. For the dual loops of motor and load position feedback, in order to obtain a high disturbance-rejection capability and robustness, the linear extended state observer (LESO) is used to online estimate and compensate the total disturbance, including parameter perturbations, unmodeled dynamics and other disturbances. On the base of the real-time compensation of the total disturbance, according to the error between input signal and feedback signal, and its differential, the linear feedback rate is decided by the controller to obtain higher control bandwidth. In addition, the control performance of the controller can be further improved by feedforward of speed. Comparative simulation and experimental results indicate that the presented active-disturbance-rejection controller has better tracking performance and robustness against the disturbances compared with the P-PI (proportional position controller and proportional-integral speed controller) control method. Moreover, the controller has the advantages of simple algorithm, clear physical meaning of control parameters, easy debugging and it is basically independent of accurate mathematical model of feed system. Active-disturbance-rejection controller, dual position loops, extended state observer, speed feedforward.