Model-Free Predictive Current Control for Switched Reluctance Motor Drives Using Current Difference Technique

This article proposes a new current control for switched reluctance motor (SRM) drives. The proposed method is based on the principles of model-free predictive control. Therefore, it eliminates the need for offline analysis to determine motor parameters and their storage in memory. The controller utilizes a current difference technique to store the measured values of current changes at each controller sampling time, during motor operation. These stored values of current difference are then mapped to their corresponding duty cycles and used to update a look-up table, aiding the decision making process of the cost function. The current difference table is memory-friendly and does not require additional current sampling, making this approach desirable for low-cost applications. An update mechanism for the table has been proposed to ensure fast update with minimal error while maintaining simplicity. Simulation results and experiments demonstrate that this fixed-switching control method can operate effectively over a wide speed range, even under the inductance saturation and induced back-EMF conditions. Additionally, the results indicate superior performance of the controller compared with different current controllers in the literature such as proportional-integral, model-based, and model-free predictive control methods in terms of low current ripple and low current root mean square error.