Active Motor Rotor Temperature Management Based on One-Node Thermal Network Model Predictive Control

In order to exploit the torque capability of electrical motors, overload operations are appreciated. However, the extent and the time of the overload operation are difficult to determine. Since the capability of overload operation is closely related to the rotor temperature for some types of motors, in this article, a novel one-node lumped-parameter thermal model is proposed and adopted by the model predictive control (MPC) to achieve the active rotor temperature management. The proposed MPC control scheme can adaptively set a current limit according to the thermal state of machines to limit the rotor temperature. Compared with the existing multinode thermal model based MPC thermal control scheme, the parameters of the proposed one-node thermal model can be easily obtained from a set of experiments, while the computational burden of the proposed MPC control was decreased significantly. The proposed control scheme was assessed by simulations and experiments on laboratory machine-drive systems and compared with a proportional-integral (PI) controller based rotor temperature control. The experimental results show that the proposed scheme achieves better control performance than the conventional PI controller based thermal control and can effectively reduce the peak rotor temperature and, hence, the thermal stress of the machine for improving its lifetime.