Three Vector Predictive Control Strategy Based on Parameter Recognition

This paper proposes a new type of composite control strategy for the problem of current loop performance due by the constant change of parameters. UKF is used to identify motor parameters in real time, solve the problem of sensitivity of traditional model prediction control to parameters, dynamically optimize the voltage vector action time allocation based on the parameter identification results, and improve the accuracy of current tracking. Simulation and experimental results show that the new composite control strategy has good speed response speed and small fluctuations in current and torque.