Fixed-switching Model-free Predictive Current Control of Switched Reluctance Motor Using Parameter Estimation

Model predictive control (MPC) is a widely known control strategy used in power electronics and electric drives fields. Its application to switched reluctance motors (SRM) is particularly attractive given MPC’s ability to handle nonlinearities while providing a high dynamic response. However, due to the dependence of MPC on the motor parameters, current tracking can be degraded as electrical parameters can change in different operating regions or when there is a mismatch in model information collected from finite element analysis. To address these issues, this paper proposes a fixed-switching model-free predictive current control (MFPC). The proposed MFPC maintains a tuned duty cycle that enables a recursive least square estimation method to estimate inductance and back-EMF values used in the predictive controller algorithm. The effectiveness of the proposed MFPC has been verified through the simulation results and compared with a proportional-integral controller. The results show improved current tracking performance and lower torque ripple using MFPC approach, in the low to medium speed range.