Continuous Control SetModel Predictive Control of Permanent Magnet Synchronous Motor

Abstract: The thesis provides design methods for developing Continuous Control SetModel Predictive Control and Linear Quadratic Regulator for Interior mounted Permanent Magnet Synchronous Motor drive. The software for these controllers are implemented in the prototype motor at China Euro Vehicle Technology (CEVT) to estimate the computation time. The performance of these controllers are compared with Field Oriented Control based on Proportional Integral regulator. The comparative analysis for these controllers are performed in Model-in the Loop environment in Simulink developed by the company. It is a back-to-back set up where the test motor is driven by a load motor. The test motor is torque controlled while load motor is speed controlled. The motor is driven by a 2-level Voltage Source Inverter using Space Vector Modulation. The switching frequency of operation is 10 kHz for all controllers. Two variants of Model Predictive Control is discussed in this project. First is the reference tracking predictive control modelled as current controller for the drive. The absence of integral action is verified. Secondly, explicit integral action in Model Predictive Control and Linear Quadratic Regulator is introduced to handle the parameter mismatches. The three controllers, namely Proportional Regulator, integral states Model Predictive Control and Linear Quadratic Regulator where tested for their steady state current and torque tracking accuracy. The steady state ripples on current and torque is compared. The transient performance of the controllers are compared for torque reference changes and load disturbances. Sensitivity of the controllers to parameter mismatches is tested. Finally, ease of tuning these controllers are discussed. The computation time for the individual controllers are obtained by testing it in a microcontroller.