An adaptive predictive current control technique for permanent magnet synchronous motors

Producing the required magnitude and shape of reference current is necessary to achieve precise torque control in motor control applications. A deadbeat predictive current controller provides very good dynamic performance. In order to achieve very accurate reference current production the machine and inverter should be modeled properly with accurate parameters. Parameter changes, especially the back emf voltages and inverter delay elements, cause deadbeat current regulator performance to deteriorate significantly. This paper addresses accurate determination of permanent magnet synchronous machine parameters dynamically on-line, taking into account the delay elements such as inverter output filters, calculation time steps of the control algorithm within the DSP, and updating the current regulation accordingly. The new algorithm is simulated and experimentally tested. The simulation and experimental results of the new current regulation technique provide very good dynamic and steady state results, validating regulator performance.