Transformer windings are usually exposed to various electrical stresses, such as the overvoltages generated in power grids. When overvoltage exceeds the winding insulation withstand level, various electrical failures can occur. For disc-type windings, overvoltages, depending on insulation condition and aging level, may lead to complete failure. Detecting the type of these failures using computerized methods can reduce the time and effort required to determine them in High-Voltage (HV) laboratories. Discriminating between electrical failures in the windings can be performed using various techniques. The Frequency Response Analysis (FRA) technique is commonly used for fault detection in transformer windings. In the present paper, the FRA technique is first used to measure the Transfer Function (TF) of a single HV disc winding under intact and faulty conditions with varying severities. Afterward, the measured TFs are arranged in a matrix and subjected to a Dimension Reduction Algorithm (DRA) to reduce their high dimensionality, extract the relevant features, and visualize them in a 2D space. Applying the DRA to the formed matrix yields two distinct clusters corresponding to disc-to-disc and turn-to-turn faults. Finally, two new unknown faults are selected as test data, and an unsupervised approach, such as k-means, is used to determine their types.
Brushless dc motor (BLDCM) drives suffer inherently from high commutation torque ripple (CTR), which introduces speed fluctuation and torque pulsation, thereby affecting the smooth operation of the drive. To overcome this difficulty, BLDCM drive topologies supported by rectifier-fed dc–dc converters have gained great research attention because of their improved controllability and energy conversion capability. The nonlinear switching characteristics of the semiconductor devices distort the input current waveform leading to increased harmonics and reduced power factor. This necessitates higher RMS current drawn from the utility grid, resulting in elevated conduction losses and localized component heating, while creating difficulties in meeting international power quality standards. In this study, a front-end Zeta converter (FEZC)-fed BLDCM drive with integrated power factor correction (PFC) stage is proposed to address these limitations. The FEZC offers a regulated dc-link voltage (DCLV) that improves current commutation and significantly mitigates the CTR, while its inherent PFC capability shapes the input current to achieve near unity power factor (UPF) and low input current harmonics, therefore ensuring compliance with standards like IEC 61000-3-2 and IEEE 519. The suggested drive topology is validated through detailed analytical modeling, simulation, and experimental investigations. The obtained findings demonstrate improved conversion efficiency, lower total harmonic distortion (THD), and notable reduction in CTR.
Grid-connected distributed energy resources (DERs) have gained significant attention due to their ability to enhance grid performance, improve resilience, and support renewable energy integration. Additionally, they reduce operational costs and provide economic benefits for both consumers and utilities. In alignment with this vision, Dubai aims to become the world’s smartest city and achieve net-zero emissions by 2050 through the integration of DERs into the power grid. To support this objective, a smart grid station (SGS) has been developed as part of Dubai’s smart city transformation. This paper focuses on the testing and validation of an existing grid-connected SGS using a real-time simulator (RTS), incorporating multiple modes of operation (MOPs). The SGS, functioning as a grid-connected microgrid, integrates solar photovoltaic (PV) systems, battery energy storage systems (BESs), wind turbines, and flexible loads. The primary objective of this work is to model the existing SGS in an RTS environment and validate its performance under various MOPs. An energy management strategy (EMS) is implemented to coordinate power exchange among the DERs and the utility grid while maintaining stable system operation under grid-connected and island conditions. To enable computationally efficient real-time execution, average converter models and established control strategies are employed for the DER interfacing converters. The proposed SGS is implemented and validated using the RT-LAB real-time simulation platform under various operating scenarios. The experimental results demonstrate stable power sharing, effective coordination among DERs, and seamless transitions between different operating modes, thereby verifying the feasibility of the proposed SGS framework for system-level real-time validation. The developed RTS platform provides a practical foundation for future research involving power hardware-in-the-loop testing, weak-grid operation, electric vehicle integration, demand response, and market-oriented energy management. The experimental results demonstrate that the ac bus and MDB voltages were maintained at approximately 226.2V and the 400V-side and MDB root-mean-square (RMS) voltages at approximately 226.9V, while the 11 kV-side phase-to-neutral RMS voltage remained approximately 6.263 kV. The validation also resulted in low-performance indices of 0.01634 for the DC-link voltage and 8.907 × 10^-6 for the ac-side RMS voltage, demonstrating close tracking of the reference values.
This paper proposes an innovative method for optimizing power grids with high renewable energy penetration by employing particle swarm optimization (PSO) to strategically allocate static synchronous compensators (STATCOMs). The approach, implemented through DIgSILENT’s programming language (DPL), aims to minimize voltage deviations and power losses while maximizing the reactive power reserve (RPR) from solar power plants to enhance grid stability and efficiency. The effectiveness of the method is tested under various system contingencies, including short-circuit faults and generation loss events. Results demonstrate that optimally placed STATCOMs significantly improve voltage profiles in the neighboring areas during disturbances. The study also evaluates different control modes for photovoltaic (PV) solar systems, with findings highlighting the importance of voltage control mode in preventing system collapse. Additionally, integrating RPR from solar power plants, particularly in combination with STATCOMs, is highly effective in maintaining a healthy voltage profile.
Electromagnetic energy harvesting-powered wireless sensors (EEHWS) are required to exhibit ultra-low power consumption, maintain operability under low-vibration conditions, and achieve autonomous fast start-up capability after long periods without ambient energy supply. However, most current EEHWS do not meet these requirements, remain inefficient under weak excitations, rely on dedicated start-up and complex gate driver circuits, even requiring up to 5 V supply voltage. Most of them do not take into account or report the time to restart after unpowered phases. To address these challenges, this paper proposes a novel AC/DC boost converter that eliminates the need for a dedicated start-up circuit by integrating a passive multi-stage voltage multiplier (VM) in the rectification stage to quickly reach the voltage level required for start-up. The number of required VM stages has been analytically and experimentally investigated in dependence on the characteristics of the electromagnetic converter and the energy profile of the output. The converter requires only a simple ancillary circuit and ensures faster start-up within 20 s instead of 43.67 s after low-energy phases because, during start-up, the VM charges the supercapacitor (SC) while supplying power to the ancillary circuit. Once the SC reaches the necessary voltage level, the system transitions to the boost phase, supplying power to the ancillary circuit and raising the output voltage for continuous operation. The novel converter has an output power of 1.2 mW with an efficiency of 72
