A look-up table-based model predictive torque control of IPMSM drives with duty cycle optimization

Model Predictive Control (MPC) is an effective method of driving motors and power converters due to its quick response, integrity, and multivariable control adaptability. A model-predictive torque control (MPTC) technique for permanent magnet synchronous motors (IPMSMs), which is computationally efficient and of low complexity, is presented in this paper. The proposed technique designs a lookup table that is independent of flux angle and torque deviation. For each control instant, this technique has to evaluate four voltage space vectors (VSV) from the lookup table, resulting in a substantial reduction in switching frequency and computational burden without compromising the performance. A maximum torque per ampere (MTPA) technique generates reference currents. The controller's complexity is minimized by eliminating the flux weighting factor from the cost function, saving time on offline weighting factor adjustments. Moreover, duty cycle optimization is performed using the mean torque control technique to minimize torque and flux ripples. The proposed method has been experimentally validated using a real-time simulator hardware in loop (HIL) with a TMS320F28335 floating-point digital signal processor on a prototype IPMSM drive. Furthermore, the proposed MPTC scheme is compared to conventional MPTC and direct torque control (DTC).