Model Prediction Hybrid Parallel Direct Speed Control of Permanent Magnet Synchronous Machines for Electric Vehicles

Abstract

Model predictive direct speed control (MP-DSC) has been widely concerned for its unique fast speed response capability and simple structure. However, the control performance of conventional MP-DSC is extremely susceptible to the influence of weight factor in the cost function and external disturbances. In this paper, a model prediction hybrid parallel direct speed control (MP-HPDSC) method is proposed. The composite prediction error cost function was decomposed into independent forms: speed error, torque error and flux error, and the optimal voltage vector (VV) was selected by optimized parallel structure to eliminate the influence of weight factor. Meanwhile, the reference value of torque is obtained by cascade predictive torque control (CPTC) combined with linear extended state observer (LESO), which can improve the anti-disturbance performance while retaining the fast speed response of system. Compared with other methods, the proposed MP-HPDSC has faster speed tracking, more stable torque output and anti-disturbance performance, which is more adaptive to the control requirements of electric vehicles system. Simulation and experiment results show the effectiveness of the proposed method.