AUT Journal of Electrical Engineering
Year:2025 | Volume:57 | Issue:1|Papers Count:15

The main advantages of the flyback converter are the simple topology and the isolation of the output from the primary, which is useful for cases where a compact design is required. The converter exhibits inherently nonlinear dynamic behavior, it is also a non-minimum phase system with a zero in the right half plane. In this paper, a combined topology of the flyback converter and conventional forward converter, or so-called “forward-flyback converter” is investigated. In addition to the fact that the forward-flyback converter has the inherent advantages of the flyback converter (isolation and simple topology), it is shown that this converter will have better dynamic behavior than the flyback converter. Obtaining the transfer function of the forward-flyback converter through the average state space equations is a difficult task. In this paper, the transfer function of this converter is obtained through a new and simple innovative method. The design of the converter to achieve minimum phase dynamics is also presented. To confirm the analysis performed, laboratory setup of the flyback and forward-flyback converters with a 50-watt power and voltage conversion of 150 to 24 volts DC is implemented, and the obtained results from the measurements show the superiority of the dynamic behavior of the forward-flyback converter compared to the conventional flyback. This work also presents two feedback control designs for the forward-flyback converter using obtained equivalent circuits and models.

The escalating development of artificial intelligence and machine learning in Industry 4.0 and cyber-physical systems has heightened security challenges for humans. In addressing this, Physical Unclonable Functions (PUFs) have emerged as a promising, lightweight solution to enhance the security of Internet of Things (IoT) devices. The imperative need for secure and low-power cryptographic devices has become evident in the IoT domain and its evolving technologies. Although IoT has enabled battery-operated devices to transmit sensitive data, it has also introduced challenges, including high power consumption and security vulnerabilities. This paper presents an exploration of the utilization of adiabatic logic with carbon nanotube field-effect transistors (CNTFETs) for the design of lightweight IoT devices aimed at addressing these challenges. The proposed computing platform and architecture circuit, employing Static Random-Access Memory (SRAM), demonstrate the potential to enhance security and energy efficiency for IoT applications. Our research showcases highly resilient CNTFET and adiabatic logic-based SRAM-PUFs, exhibiting an ultra-low start-up power of 1.8 nw. The PUF metrics, including uniformity, reliability, and uniqueness, are 46.10%, 88.47%, and 48.84%, respectively, across a 150% process variation. In this paper, we conduct circuit simulations using 32nm CNTFET technology in HSpice to scrutinize the impact of threshold voltage fluctuations. Further post-processing procedures are executed using MATLAB software.

Since Huygens structures with phase shift lead to beam rotation, this manuscript discusses a phase shifter structure for the Huygens surface based on Tapered Slot Transition in the frequency range of 9.95-10.2 GHz using varactor diodes. By inducing an electric field along the tracks and placing diodes, the phase can be changed up to 200 degrees. With unit cells of 0.2  -that this size is the smallest possible size for beam rotation structures - and a mega cell of 4 unit cells, incident angles of 0-15 degrees can be converted to transmitted angles of 10-45 degrees. The unit cell has been fabricated and it is shown that the measurement results are in good agreement with the simulation one.

In various industries, the coordinated movements of mobile robots in triangular formations hold promise for enhancing efficiency and safety. This study investigates trajectory tracking and formation control using two distinct methodologies: the PID controller and the Fuzzy Logic Controller (FLC). Under ideal conditions, both controllers exhibit precise navigation and formation maintenance. Notably, the leader robot has a simulated virtual sensor for obstacle avoidance. The followers emulate the leader’s path using the selected controller methodology. However, when exposed to external disturbances, modeled as sinusoidal waves, the FLC, with its superior adaptability and resilience, demonstrates its potential as a robust solution for real-world applications susceptible to disturbances. This research emphasizes the pivotal role of controller selection in practical scenarios and reiterates the FLC’s potential, instilling confidence in its effectiveness.

With the proliferation of smart mobile devices, there is an ever-increasing demand for multimedia content. To avoid congestion in backhaul links, mobile edge caching is a promising solution that can reduce delivery delays and improve users' quality of experience. In this regard, the requested content can be downloaded from a nearby small cell access point (also called helper) instead of a base station with a lower delay. In this paper, we address the problem of finding the optimal cache data placement to minimize the total delivery delay. We suppose the users are flexible in the sense that they request a set of multiple files from the library with a unique feature and are satisfied if any file within the requested set is received. Moreover, in the system model, the interference and the mobility of users are considered. More precisely, the effect of interference from other helpers is incorporated in calculating the delivery delay, and a random waypoint model is exploited to address the mobility of users within the network. Because of the complexity of the problem, finding the optimal solution is NP-hard. We prove that the problem is in the form of maximizing a monotone submodular function subject to matroid constraints. We exploit this property to provide an efficient approximate solution (i.e., a greedy algorithm) that is guaranteed to perform within a constant of  as well as the optimal solution. Simulation results validate the efficiency of our proposed algorithm.

In this extended study, the focus is on advancing the generation of synthetic distribution grids (SDGs) through the introduction of a new algorithm based on the Barabási-Albert random graph model. The initial use of the Erdős model to create SDGs revealed limitations in size and structural adjustability beyond the number of vertices. To address these limitations and push the research forward, the new algorithm utilizes the Barabási-Albert model to provide more control over the structural features of the generated graphs through the introduction of a novel tuning parameter known as the “richness index”. The effectiveness of both algorithms in producing SDGs of various sizes is demonstrated by generating SDGs with different sizes, confirming their ability to mimic synthetic radial distribution grids successfully. Additionally, a detailed examination of degree-based parameters and Pearson coefficients for SDGs of sizes from 20 to 1000 uncovers significant patterns. Furthermore, the proposed algorithm is examined in the terms of the variation of richness index in branching rate and μ-PMU placement, confirming the scale-free characteristic of the method. A comparison of the Erdős and Barabási-Albert models shows variations in maximum degree values, branching rates, and mixing patterns. The original Barabási-Albert model tends to have nodes with higher degrees and increased branching rates, which can be adjusted by the richness index. These findings emphasize the ability of the Barabási-Albert model to generate scale-free SDGs with diverse structures by fine-tuning the richness index.

This paper introduces a new high step-down converter that achieves zero voltage switching in both primary and auxiliary switches without relying on coupled inductors, thus ensuring low ripple in input and output currents. The auxiliary circuit is optimized with the fewest possible components, facilitating zero voltage switching for the main and auxiliary switches and solving the reverse recovery issue of freewheeling diodes through zero current switching conditions. Additionally, this design reduces voltage stress on the switches, allowing the use of MOSFETs with lower drain-source resistance. A key advantage of this converter is that it avoids using coupled inductors, which eliminates issues related to leakage inductance and potential increases in the converter's size and weight. This enhances the converter's efficiency and practicality. The auxiliary circuit design can also be extended to multiple phases, making it adaptable for various applications. The control mechanism is simple because the auxiliary switch operates in complementary with the main switch, simplifying the overall circuit design and integration. The proposed converter's design and operational principles have been validated through PSpice simulations, and a 90W prototype has been constructed. Experimental results demonstrate an impressive 95 percent efficiency at full load.

Today the application of Distributed Energy Resources (DERs) in Distribution System expansion planning (DNEP) problems is more crucial than before. Despite of advantages, the presence of the resources considering Renewable Energy Sources (RESs) and Dispatchable Generation (DG) units in the distribution System expansion planning problems, brings more challenges, especially in reliability characteristics. This paper proposes a new distribution network expansion planning model embedded with a novel reliability assessment approach for Active Distribution Networks (ADNs). The proposed method aims to determine the optimal location and capacity of the new generation and distribution assets, responsible for providing power, in both the normal operation and contingency conditions. The load forecast significantly affects the results of the distribution network expansion planning. The K-means clustering method is used to address the uncertainty of load growth in the planning horizon which is coordinated with a Mixed Integer Linear Programming (MILP) optimization model. The proposed model is applied to the IEEE 33 bus test case, to guarantee its technical and economical effectiveness. The results verify that this model is cost-effective and can increase the robustness of the distribution network compared with recent similar works.

The reduction of fossil fuel reserves, advancements in science and technology, increased network loading, the emergence of new energy sources and loads, etc., have contributed to the rise of smart grids. Within smart grids, distributed generation resources play a pivotal role in meeting the network's power requirements. Among these resources, renewable energies and electric vehicles are notable examples. In this context, the presence of electric vehicles on smart grids has led to numerous opportunities and challenges, underscoring the need for effective management of these vehicles. Among these concepts, a relatively new one known as the “Electric Vehicle Aggregator” is introduced. This aggregator provides the opportunity to participate in the demand-side management of energy networks by managing the scheduling of electric vehicle charging and discharging. In this paper, an attempt has been made to reduce the energy received from the grid by using the solar microgrid and considering the ability to connect to the upstream grid, and designing a new aggregator to maximize the profit of the owner of the aggregator. The proposed model has been designed, implemented, and tested over a 25-year time period using Homer software. The simulation results also show that using the proposed model despite considering the initial investment of the solar microgrid in the target functions, improvement and increase in the profit of the aggregator owner. The cost of the aggregator in the proposed method of this paper is 24.48$, while the same cost in the method used in the main reference is 29.62$.

In today’s industrial world, it is indispensable to offer a correlated equilibrium between the reconstruction costs of damaged electric motors and desirable of performance characteristics so as to exploit available primary materials. This paper presents a cost-oriented scheme to choose an effective insulation method in reconstructing the medium- and high-voltage stator coils in both vacuum pressure impregnation and resin-rich methods. The insulations required in coil reconstruction include three sections: strand, group, and main body insulation. These insulations have specific electrical characteristics and dimensions that limit their use in a specific part of the coil. On the other hand, sometimes these insulations are not available or it is difficult to get them. Because of this, their prices are constantly changing, and it is very difficult to estimate the cost of materials relatively accurately. In this paper, various designs are produced by combining available insulation from each section. Then, mathematical models will be extracted to check the functional characteristics and costs of renovation materials in the plans. Ultimately, the best reconstruction design, changes in performance characteristics compared to insulations with different specifications is selected and the sensitivity of the cost of the designs to the cost of insulation materials and the effects caused by Remove group insulation is checked. For validation, the results associated with the reconstruction calculations of a damaged realistic 6-kV 535-kW electric motor are presented.

Traditionally, diagnosis of bearing faults involves analyzing the frequency spectra of monitored signals, like vibration and stator current, using various signal processing techniques. However, signal-based methods for fault diagnosis often produce false alarms due to changes in load and voltage imbalances in the motor's input. Furthermore, these methods have limited performance in detecting faults at early stages and readjusting based on speed, load, and voltage levels. To overcome these challenges, this paper proposes a model-based approach for bearing fault diagnosis utilizing the Luenberger observer. The suggested model-based method compares the real behavior of the system with the estimated behavior of its nominal model, eliminating non-fault-related factors that have similar effects on both the system and its mathematical model. The efficiency of the suggested model-based bearing fault diagnosis method is validated by comparing simulation and experimental results obtained from the proposed model-based method with a recent signal-based method. The proposed method introduces a novel application of the Luenberger observer for fault detection in induction motors, offering a simple and efficient approach to diagnosing bearing faults. It uniquely distinguishes mechanical faults without direct electrical signal correlation and incorporates a systematic noise cancellation technique, enhancing robustness and accuracy under varying loads.

Enhancing the efficiency and environmental compatibility of hybrid systems through renewable energy sources is a highly motivating concept. Two tourist destinations, one Sarab Gian region in Nahavand city of Iran and the other in Zurich city in Switzerland, were analyzed. Using HOMER Pro software, photovoltaic panels (PVs), wind turbines (WTs), battery energy storage systems (BESS), and diesel generators (DGs) were evaluated. Sensitivity factors such as varying fossil fuel costs, fuel supply limitations, inflation rates, discount rates, carbon dioxide penalties, and capacity shortages were taken into account. The findings indicate that as fuel prices and emission penalties rise, renewable resources become more cost-effective. Comparing the lowest net present cost (NPC) in Switzerland to Iran, which is also influenced by fuel prices, 180% increase was observed. Furthermore, due to the impact of fuel prices and the optimal capacity of PVs, we observed a 5.3% increase in total operational costs in scenario Sarab Gian (CA) compared to Zurich (CB). Additionally, Switzerland benefits from lower inflation and discount rates, leading to a 19.3% reduction in NPC and a 41% decrease in the cost of energy (COE) compared to Iran. This study emphasizes the economic feasibility of renewable energy in hybrid systems, particularly in regions with high fossil fuel costs and strict emission regulations.

Distribution locational marginal pricing (DLMP) is an efficient approach to optimize the pricing of distribution systems. This paper focuses on DLMP to minimize losses within the distribution network. This approach can be strategically manipulated to adjust the profits for distributed generation (DG) owners and the distribution company. Furthermore, the paper employs the information gap decision theory (IGDT) method scenarios to model the uncertainty surrounding electricity market prices. By incorporating the risk-averse (RA) scenario, network operators can discern RA solutions and optimal outcomes derived from the algorithm. On the other hand, the risk-tolerance (RT) scenario helps identify riskier solutions, enabling appropriate decision-making based on whether the solutions are RA or risky in nature. To further enhance the quality of outcomes, this paper combines IGDT scenarios with the cheetah hunter optimization (CHO) algorithm to ensure the obtained results are both optimal and accurate. The proposed method’s performance is evaluated through simulations conducted on a 69-bus IEEE power network using the MATLAB software environment. The results obtained from this approach demonstrate its superior accuracy when compared to previous methodologies.

Current-based methods for bearing fault diagnosis primarily rely on analyzing the current signal, leading to challenges in detecting fault frequencies due to their low magnitude amid the noise in the current spectrum. This issue intensifies for weak bearing faults in their early stages. The presence of noise components increases the risk of false alarms, as fault characteristics are often obscured in the raw current spectral analysis. To address this, effective bearing fault diagnosis necessitates the reduction of noise components. This paper presents a novel noise cancellation method that enhances the estimation of bearing fault signals in induction motors by utilizing the deviation of instantaneous frequency in synchronized motor voltage and current signals. The proposed method efficiently diagnoses bearing fault characteristic frequencies during spectral analysis. Simulation and experimental results substantiate the effectiveness of this approach in detecting outer/inner raceway and ball-bearing faults.

Power-electronic converters play a crucial role in the functioning of microgrids. However, these converters, characterized by their low inertia, present a significant challenge to maintaining a consistent frequency in islanded microgrids. To address this issue, an innovative concept known as virtual inertia control has emerged as a promising solution for enhancing frequency stability in islanded microgrids. The virtual inertia control system does not perform well against disturbances and uncertainty related to microgrid parameters. Therefore, to overcome these problems, it needs a suitable controller in its structure. In this paper, a linear quadratic regulator mode feedback controller based on deep learning is proposed to improve the performance of virtual inertia control in an islanded microgrid against disturbances and uncertainties in the system. The linear quadratic regulator controller uses measurements of system states and the integration of a deep network increases the accuracy and dynamic response of the feedback controller. This allows for fine-tuning of the control response, which exhibits significant robustness against uncertainty in system parameters and disturbances. To evaluate its effectiveness and compare it against alternative control approaches, comprehensive assessments have been conducted across multiple scenarios. The results indicate that the proposed method in the field of virtual inertia control surpasses previous approaches.