Model Predictive Torque Control Of Permanent Magnet Synchronous Motor Based On Optimal Duty Cycle

Permanent magnet synchronous motor (PMSM) has been widely used in aerospace,numerically-controlled machine tools and other high-performance occasions due to its simple structure and high efficiency. Because of its simple principle and easy to deal with the nonlinear constraint conditions, the model predictive control has strong applicability. Model predictive torque control (MPTC) strategy is the application of model predictive control in direct torque control. There are only six effective voltage vectors and a zero voltage vector can be selected in two level inverter, and only one voltage vector can be applied in a control cycle for traditional MPTC, so there are large torque and flux ripples in the traditional MPTC. Torque and flux ripples can be reduced by the duty cycle MPTC. However, in this method the voltage vector is selected firstly, and then the duty cycle is calculated, so the voltage vector can not be globally optimal. To solve this problem, based on the analysis of voltage vector amplitude and torque and flux increment, an optimal duty cycle model predictive torque control strategy is proposed in this paper. The duty cycle of each voltage vector need be calculated firstly which make torque and flux error minimum at next instant, and then the combination of the optimal voltage vector and the duty cycle can be selected by the cost function. Both torque and flux are simultaneously considered during the calculation of duty cycle.In order to verify the feasibility and effectiveness of the proposed control strategy, the traditional MPTC, the duty cycle MPTC and the optimal duty cycle MPTC are simulated in MALTAB/Simulink. The experiments of three strategies were carried out on a two level inverter. Experimental results show that the duty cycle MPTC and the optimal duty cycle MPTC can reduce torque and flux linkage ripples and inherited the fast dynamic response of the MPTC, and the optimal duty cycle MPTC can reduce torque and flux linkage ripples more effectively.