Direct Instantaneous Torque Control of Switched Reluctance Motors With an Asymmetric Converter Using Adaptive Switching Angles

Direct instantaneous torque control (DITC) is a method for torque control in the switched reluctance motor (SRM). In conventional DITC, due to the fixed turn‐on and turn‐off angles, the presence of negative torque and significant torque ripple is inevitable. This paper proposes an improved DITC strategy for SRM drives based on adaptive turn‐on and turn‐off angle regulation to reduce torque ripple and improve transient response. Unlike conventional hybrid switching‐angle optimization schemes that jointly tune both angles using a unified search strategy, the proposed method assigns distinct functional roles to the two stages: the turn‐off angle is estimated by artificial neural network (ANN) as a data‐driven offline mapping, whereas the turn‐on angle is corrected online by a self‐tuning fuzzy‐PID controller based on torque error dynamics. The proposed approach focuses on a hybrid control system whose ultimate goal is to dynamically adjust the turn‐on and turn‐off angles of the phases to achieve a smoother output torque profile closer to the reference torque. This system, by utilizing a dual control structure, attempts to manage the dynamic uncertainties and nonlinearities of the SRM magnetic characteristics. First, an ANN is proposed to optimize the turn‐off angle ( θ _off ) under various operating conditions to prevent the generation of negative motor torque. Then, the turn‐on angle ( θ _on ) is adjusted via a self‐tuning fuzzy PID controller. The estimated torque average is compared with the reference torque. The conventional DITC hysteresis loop still relies on instantaneous torque control. The average torque is used only in the supervisory turn‐on‐angle adjustment because it better reflects the net torque‐producing capability over one electrical interval and avoids unnecessary angle chattering caused by instantaneous torque oscillations. The optimal reduction value for the turn‐on angle is calculated using a self‐tuning fuzzy PID controller and subtracted from the θ _on . The θ _on compensation is activated when the difference between the average torque and the reference torque exceeds a threshold value. The proposed switching‐angle adjustment primarily targets torque ripple reduction and torque‐response improvement, while also mitigating negative torque and the associated estimated copper loss. In addition, the proposed strategy does not reduce the turn‐on angle at low speeds, which consequently diminishes copper losses. Simulation results show a torque ripple reduction of up to 55% and a rise time improvement of up to 60%, confirming the effectiveness of the proposed method for high‐speed applications. To validate the proposed method, experimental results will be presented.