Contents I. Introduction
Switched Reluctance Motors (SRMs) are cost-effective, durable, fault-tolerant, and independent from rare earth magnets. Hence, they serve as a good alternative to induction and permanent magnet motors in automotive and space applications [1], [2]. However, their double salient construction and discrete commutation cause torque ripple, vibrations, and acoustic noise. Controllers with hysteresis are typically used for current regulation and reference tracking in SRMs. However, significant current ripples are observed in these cases [3]. Model predictive control (MPC) improves non-linear system performance by using a system model to predict machine behavior. MPC was first used in the chemical industry to solve problems like long-time constants, high computing demands, and limited computational processors. Advancements in microprocessor and semiconductor technology have made MPC implementation in motor drive systems easier [4]. Predictive control implementation uses numerous methods, including FCS-MPC, CCS-MPC, and Dead beat control. A modulator generates switching states from the predictive controller’s continuous output in the Continuous Control Set Model Predictive Control (CCS-MPC). FCS-MPC uses fewer switching states for power converters, which is an advantage. A key advantage of FCS-MPC is the absence of an intermediary modulation stage [5]. The FCS-MPC method is popular for SRM motors due to its low computational burden and ease of implementation. MPC is commonly used to regulate induction and permanent magnet synchronous motors. The use of MPC in SRM is still in its early stages and not as mature as in other motor types [6]. The typical MPC method for SRM forecasts flux, current, and torque using the SRM mathematical model. The Forward-Euler method discretizes the mathematical model. The discretized model often has large errors and global instability. This paper introduces different SRM discretization methods to improve flux, current, and torque estimation. FCS-MPC accuracy depends on model discretization precision. This study compares discretization options for the mathematical model of SRM to the traditional approach. The paper’s subsequent sections follow this structure: Section II of this research models the FCS-MPC SRM. SRM discretization approaches are reviewed in Section III. Section IV discusses the stability of Forward Euler and Rungekutta methods of discretization, also a thorough comparison is done for different discretization methods. This study concludes in Section V
