Model Predictive Torque Control of Switched Reluctance Machines With Torque Sharing Function and PWM Control Signals Based on Linear Polynomial Fitting
Lefei Ge, Jixuan Guo, Chao Gong, Guoqiang Zhang, Xiaofeng Ding, Shoujun Song
Abstract
In the combination of torque sharing function (TSF) and model predictive torque control (MPTC), the finite set prediction with only three evaluated switching states can not fully exploit the capability of MPTC. This letter proposes a novel MPTC with the TSF and pulse width modulation (PWM) control signals based on linear polynomial fitting for the torque ripple suppression of the switched reluctance machine (SRM). The sinusoidal TSF is used to distribute the reference torque to each phase. To fully utilize the flexibility of PWM control signals, the linear polynomial fitting method is adopted to expand the predictive data obtained by the prediction model, which avoids the redundant lookup table calculation. Then, the optimal duty cycle corresponding to the minimum torque tracking error is found and applied. The experimental results show that the proposed MPTC has good reference torque tracking capability and torque ripple suppression performance.