Threatened by weather disasters and operational uncertainties, power systems require resilient and cost-effective decision making to ensure security. This article proposes a novel deep reinforcement learning algorithm, namely defender–attacker soft actor-critic (DA-SAC), designed for contingency-constrained optimal power flow underN–ksecurity criteria. A two-agent Markov decision process is formulated, where the defender learns robust control actions and the attacker identifies worst-case contingencies. The core soft actor-critic algorithm is enhanced by integrating constraint violation levels into the reward function and employing a two-timescale learning scheme to improve feasibility and stability. The proposed method is validated on the IEEE 30-bus and 118-bus systems. Simulation results show that DA-SAC significantly reduces unserved energy, load shedding, and constraint violations, outperforming conventional and deep-reinforcement-learning-based benchmarks underN–1,N–2, andN–3scenarios. These results demonstrate that DA-SAC offers a fast, resilient, and practical solution for real-time power system operation under severe contingencies.

This paper proposes a novel cascaded multilevel inverter (CMLI) structure based on switch capacitor (SC) tech-nology, which is capable of achieving up to 17 levels. The inverter is constructed by cascading two SC-based 9-level MLI basic modules, resulting in a 17-level SC-CMLI structure. The proposed inverter employs a reduced number of power switches, floating capacitors, gate drives, and dc sources, making it easy to implement. The Nearest level Control PWM technique is used to generate the switching pulses for the proposed 17-level CMLI. The proposed inverter is simulated under different modulation indices and dynamic load change conditions. The output voltage waveforms have significantly less total harmonic distortion (THD) content of 5.18%. The theoretical analysis of the proposed inverter is verified using MATLAB/Simulink software. Overall, the results demonstrate that the proposed 17-level SC-CMLI is an efficient and practical solution for high-power applications. This topology can be extended to higher voltage levels.

This research proposes a sensorless Field Oriented Control (FOC) for Permanent Magnet Brushless DC Motor (PMBLDCM), employing a Continuous Fast Terminal (CFT) Sliding Mode Controller (SMC) which incorporates fault-tolerant operation for low cost electric vehicle. The approach presented here provides a robust quantitative design and execution of the speed controller, and integrates an estimation mechanism for total disturbances, including the effects of iron loss on electromagnetic torque. Initially, the rotor speed dynamics of the PMBLDCM drive system are studied, considering the lumped disruption which encapsulates interference, parametric complexities, and nonlinear dynamics. In this context, the applied Sliding Mode Observer (SMO) controller works to regulate real-time rotor speed and counteract lumped disruptions in rotor speed mechanics, thus ensuring fault-tolerant operation. The FOC scheme is then further enhanced to improve the generated torque while significantly minimizing ripple content. The proposed methodology eases the zero torque-pulsation restriction, thereby broadening the torque operation range. Furthermore, the technique can mitigate torque ripples throughout the complete torque operating spectrum for each reference value of torque. Ultimately, the introduction of the sliding mode control strategy increases the robustness of the PMBLDCM system, ensuring efficient and reliable operation under various disturbance conditions.

In the electric vehicle industry, multilevel inverters (MLIs) are widely used for power conversion in high-power-medium-voltage propulsion drives. The five-level ANPC topology with voltage boosting capability is a promising MLI topology for single-stage solar photovoltaic power conversion. The switching pulses for the MLIs are generated using various PWM techniques. The performance of the five-level boosted ANPC inverter is compared using four distinct unipolar PWM techniques: unipolar sine PWM, unipolar 60◦PWM, unipolar third harmonic injection (THI) PWM, and unipolar zero sequence injection (ZSI) PWM. The performance of these PWM schemes are tested on real-time OPAL-RT platform. Under unipolar THI PWM, the 5L-boosted ANPC outperforms in terms of %THD reduction and higher fundamental output voltage magnitude. Furthermore, the ZSI PWM balances the neutral point potential efficiently.

In this paper, a novel stator ground fault detection method based on wavelet theory is tested on a lab scale synchronous generator. The lab scale setup is used to mimic the real power generation system and to simulate the fault characteristics particularly at or near the generator neutral of large industrial generators. The experimental results prove the capability of the wavelet algorithm in synchronous generators stator winding ground fault detection.