A Comparison of Finite Control Set and Continuous Control Set Model Predictive Control Schemes for Speed Control of Induction Motors

作者：Ahmed, Abdelsalam A.; Koh, Byung Kwon; Lee, Young Il^*

来源：IEEE Transactions on Industrial Informatics, 2018, 14(4): 1334-1346.

DOI：10.1109/TII.2017.2758393

• Continuous control set-model predictive control (CCS-MPC)
• field weakening control
• finite control set MPC
• induction motors
• Kalman filter (KF) estimator
• speed control
• torque constraint

This paper presents a comparative study of two predictive speed control schemes for induction machine (IM) in terms of their design and performance. The first control scheme is finite control set-model predictive control (FCS-MPC) with modulation control and the second control scheme is continuous control set-model predictive control (CCS-MPC) with space vector-pulse width modulation. The two schemes adopt the cascaded control approach, which consists of an inner MPC loop for torque control and outer MPC loop for speed control using two individual cost functions. The outer MPC produces the required torque to drive the IM at the reference speed while the reference torque is taken as the input of the inner MPC, which in turn generates control signals for the inverter. The control states of the two MPCs are constrained with the maximum limits of the drive system. The state feedback is achieved with a standard Kalman filter, which estimates the nonmeasured load torque. For a fair comparison, both approaches are applied to the same IM at the same operational circumstances. The control approaches are implemented and validated in an experimental environment using the same sampling frequency on the same test bench (3.7 kW IM drive). The behavior of the control approaches is assessed by applying reference and disturbance steps to the system in different operational modes. Comparison of the predictive schemes leads to the conclusion that the both MPC approaches achieve similar performances. However, the CCS-MPC scheme has a smaller current ripple and is of low computational complexity. The computing duration is not very different for the three tested schemes. CCS-MPC can cope with a less powerful DSP than for FCS.

• 出版日期2018-4