Antirollback Control for Gearless Elevator Traction Machines Adopting Offset-Free Model Predictive Control Strategy

Direct-drive permanent magnet traction systems have become the trend of modern elevators. Elevator traction machines tend to use low-resolution incremental encoders as position and speed feedback devices because of their low cost. To prevent the occurrence of elevator car rollback for the gearless elevator installed with an incremental encoder with general resolution, a weight-transducerless starting torque strategy based on offset-free model predictive control is proposed. Under the condition of low-resolution incremental encoders, this control strategy can achieve a fast dynamic response and no mechanical vibration during the mechanical brake releasing. The electromagnetic torque generated by the machine can quickly balance the uncertain load torque to minimize the sliding distance of the elevator car, which contributes in achieving superior riding comfort. In order to overcome the mismatch of the predictive model caused by the uncertainty and nonlinearity of the mechanical model of the traction system, as well as intense disturbance, a model corrector is added to adjust the mismatched predictive model. The steady-state speed error can be eliminated by adding the equivalent disturbance estimator to the predictive model during the zero-servo operation at elevator startup. Finally, both simulation and experimental results are provided to verify that the proposed control strategy can achieve shorter sliding distance, smaller sliding speed, and faster dynamic response.