Instantaneous direct torque control (DTC) for Switched Reluctance Motors (SRMs) typically employs a hysteresis regulator with a variable chopping frequency, which inherently results in significant torque ripple. To address this issue, this paper proposes a sliding mode-based DTC strategy that enhances robustness and effectively reduces torque ripple. The nonlinear switching term in the sliding mode controller actively rejects both internal and external disturbances, enabling accurate torque tracking. However, achieving high robustness through large switching gains often leads to severe chattering, which increases torque ripple and reduces system efficiency. To overcome this, a Lyapunov-guided fuzzy adaptation mechanism is introduced to estimate the switching gain online. This adaptive approach mitigates chattering while preserving the disturbance rejection capability of the controller. Simulation results demonstrate that the proposed controller achieves precise torque tracking and significantly lower torque ripple compared to instantaneous DTC.

As the increased penetration of wind power generation, the interactions between the power electronic converters and the utility grid have received attention. The inverter-based resources usually utilize the small-signal or impedance model to conduct dynamic stability assessment. However, wind power generators that fully consider electromechanical and electromagnetic transients generally have high-order models. It makes long-term stability analysis and numerical simulation of large-scale wind farms challenging and time-consuming. This paper proposes a reduced-order dynamic model of permanent magnet synchronous generator based on singular perturbation theory. The order reduction dynamic model shows the same dominant characteristics and similar time-domain simulation results compared to the full-order model, while the computational time is significantly reduced.

This paper proposes a sensorless control strategy for non-sinusoidal permanent magnet synchronous machines (PMSM) integrating a Super-Twisting Sliding Mode Observer (STA-SMO) with a Dual Second-Order Generalized Integrator and Positive Sequence Detector (DSOGI+PSD). The non-sinusoidal back-electromotive force (BEMF) in such machines introduces harmonic distortion, which degrade conventional estimation methods. The STA-SMO robustly estimates the BEMF with minimized chattering through finite-time convergence, while the DSOGI+PSD adaptively isolates the fundamental component by suppressing harmonics. A Phase-Locked Loop (PLL), synchronized with the filtered BEMF, dynamically extracts rotor position and speed. Simulation results demonstrate precise speed estimation and oscillation-free position tracking under steady-state and dynamic conditions. The proposed scheme effectively mitigates harmonic interference, enabling reliable sensorless control for PMSMs with non-sinusoidal BEMF.

Accurate estimation of the state of charge (SOC) and capacity of lithium-ion batteries (LIBs) is essential for ensuring the reliability of energy storage systems. Conventional model-based methods often suffer from the strong coupling between these states, complicating estimator design and reducing robustness. This paper proposes a novel estimation framework that achieves precise and decoupled estimation of SOC and capacity. A battery decoupled iterative modelling (DIM) is developed, in which a dedicated cost function is designed to iteratively estimate the optimal SOC and separately identify the battery capacity. To further enhance robustness against measurement noise, the Kalman Filter (KF) is integrated into the estimation loop, providing active noise suppression, and improving the overall reliability of the estimation process. Experimental validations under a variety of dynamic operating conditions demonstrate that the proposed framework significantly improves estimation accuracy while effectively mitigating the impact of sensor noise.

In recent years, the development of Lithium-ion battery technology has promoted the utilization of battery management systems (BMS). In the BMS, developing a reliable state of charge (SOC) estimation method is essential for managing and controlling batteries. Most of the research in this area has focused on the design of model-based observers. This paper proposes the self-adaptive super twisting sliding mode observer (SASTSMO) and improved particle swarm optimization (IPSO) algorithm to estimate SOC, which could effectively suppress the system chattering. The SASTSMO can estimate total disturbances in battery model without chattering issues and then compensate them to enhance the accuracy of SOC. Furthermore, the adaptive algorithm can adjust the gain of the observer according to the error automatically, which improves the robustness of the algorithm. Finally, the feasibility of the proposed method is verified under different test conditions.

Synchronous optimal pulse-width modulation (SOPWM) is an effective solution to reduce switching frequency without increasing the current distortion. This method is widely used along with stator flux tracking in traction applications of induction motors. However, stator flux tacking is not an ideal control technique for permanent magnet synchronous motors (PMSMs). This paper studies SOPWM with stator current tracking for PMSM traction applications. Stator current tracking with SOPWM suffers poor steady-state performance when the PMSM operates around discontinuity of the switching angles due to the voltage disturbance generated from current harmonics. This problem is usually solved by using continuous switching angles. However, this approach brings more overall current distortion due to sub-optimal switching angles. To address this issue, a novel hysteresis control-based SOPWM scheme with continuous switching angles is proposed. Preliminary simulation results indicate that the proposed method can offer low current distortion compared to conventional and continuous SOPWMs.

The vibration/noise and torque ripple are two inherent issues for switched reluctance motors (SRMs), one of the effective methods to address these two issues is flattening both the total radial force and total torque. In this paper, an optimized quadratic-sharing-function-based radial force and torque simultaneous flattening method is proposed for SRMs. Firstly, the features of the radial force and torque characteristics are analyzed without considering saturation. Then, with these unique and specific features, four regions are defined to develop the simultaneous radial force and torque control method based on the linear sharing functions. To incorporate the saturation effects in the proposed method, the quadratic sharing functions are adopted. The proper parameters of the quadratic sharing functions and the region definition at different operating conditions are obtained through the genetic algorithm (GA) optimization. The switching angles and reference current of the conventional current chopping control (CCC) method are also optimized by GA for fair comparison purpose. Extensive simulation results are presented in this paper, and these results prove the effectiveness and superiority of the proposed method regarding the simultaneous radial force and torque control.