فهرست مطالب

Majlesi Journal of Electrical Engineering
Volume:18 Issue: 1, Mar 2024

  • تاریخ انتشار: 1403/03/07
  • تعداد عناوین: 26
  • Fariborz Parandin *, Alireza Mohamadi Pages 1-8
    In this article, a two-dimensional photonic crystal-based XOR gate is designed and simulated. For this purpose, an initial two-dimensional photonic crystal structure is chosen and waveguides are created for inputs and outputs. Then, defect rods are selected so that the obtained outputs are approximately consistent with the XOR gate truth table. After that, we will be looking for the best output so that the highest optical power is created for logic 1 and the lowest power in the logical state of 0. For this reason, the simulation is performed for four defect rods and the output is obtained for different the radius of the rods. Then, using the K-Nearest Neighbors algorithm, which is a machine learning algorithm, the best output for logic 0 and also for logic 1 is obtained. The results show that the designed logic gate has high output power in logic 1 and very low power in the logic 0 state.
    Keywords: Optical XOR gate, photonic bandgap, Photonic Crystal, K-Nearest Neighbors
  • Somayeh Arab Najafabadi, Sara Nazari *, Nafiseh Osati Eraghi Pages 9-19
    Kingship verification is a process that two or more people has a family relation such as father and son or other family relation. Numerous studies have been presented to investigate the relationship between people. Kingship verification can be done based on image of face. Most of the methods presented on face images work well on face data sets recorded under controlled conditions. However, due to the complex nature of environments, rapidly and accurately examining human kinship in real-world unrestricted or wild-type scenarios is still a challenging research. In this paper, in order to overcome the aforementioned challenges, an efficient and new method is presented. In the proposed method, a method is used to launch the operation to create a map. The created feature map is stable against deformation, transition, scaling, direction and Dilation in wild images. Group-Face and TSKinFace databases are used for simulation. In order to evaluate the evaluation of the proposed method, average recall of 94.1, precision 94.6, accuracy 94.7, specificity 93.8, and finally F_Measure 95.0 were used. The superiority of the proposed method in all comparisons shows the effectiveness of the proposed method in diagnosing kinship.
    Keywords: kinship, deep learning, Dispersion wavelet, Convolution Neural Network
  • Amirhosein Asilian, S. Mohammadali Zanjani * Pages 21-32
    The unique properties of carbon monoxide and its high combustibility have led to the creation of various ‎sensors, such as electrochemical sensors and different circuits, to read its output. In this article, a deflection-type ‎Wheatstone bridge is used to measure changes in the sensor resistance, and the output voltage is connected to a 12-‎bit analog-to-digital converter through an adjustable precision amplifier. Next, a new method is proposed for self-calibrating the CO sensor. The Levenberg-Marquardt backpropagation algorithm (LMBP) is utilized in the Artificial ‎Neural Network model to minimize the Mean Squared Error (MSE) and identify the most suitable parameters in the ‎proposed method.‎‏ ‏The model under consideration has been developed and trained using real-time data.‎‏ ‏Based on ‎the experimental and evaluation outcomes, it can be concluded that the suggested model has an MSE value of ‎‎0.28249 and an R2 coefficient of determination of 0.99992, indicating high accuracy and precision. The proposed ‎sensor and calibration method have potential applications in various applications, including industrial and domestic ‎environments where CO monitoring is necessary.‎
    Keywords: Electrochemical sensor, CO monitoring, Levenberg-Marquardt backpropagation algorithm, Mean squared error, ‎Training-Validation, Testing (TVT), coefficient of determination
  • Samaneh Sadat Seyed Abolghasemi, Mehran Emadi *, Mohsen Karimi Pages 33-44
    Ultrasound images and ultrasound imaging method is an effective method in examining the challenges, problems and diseases related to the breast in women. The contrast of these images is generally very weak, however, the tumor tissue and calcium grains are evident in it. Methods based on image processing are widely used in breast tumor diagnosis and classification. In this article, a method based on pattern recognition is presented in order to detect the type of tumor. GLCM-based features are extracted from the target area, and Gabor and texture features. Then it is reduced with the help of dimension reduction methods based on principal component analysis. Finally, with the help of the improved classification of Ada KKN with ELM, they are grouped into three categories. Evaluation criteria such as Accuracy (98.81%),  sensitivity (91.51%) and specificity (94.54%) compared to other similar methods show the superiority of the proposed method.
    Keywords: Ultrasound images, Tumor, Breast, pattern recognition, Dimension Reduction
  • Mohammad Sabzevari, Masoudreza Aghabozorgi Sahaf * Pages 45-53
    Currently, satellite navigation systems on cars provide a promising means of making these vehicles unmanned in the future. There is a common problem with these systems due to the unavailability of satellite signals in tunnels, forests, and noisy areas. One of the methods used to solve this problem is by using inertial navigation systems as an auxiliary system. This system works by modeling errors and correcting them when GNSS signals are absent. A number of methods are available for error modeling such as Kalman filters, neural networks, and so on, each of which has its own advantages and disadvantages. The error of inertial navigation is modeled using LSTM deep neural networks in this article. In this neural network, the relationship between current and past data is modeled as long as the GNSS satellite signal is available to improve the output position of the inertial navigation system when the GNSS satellite signal can no longer be received. The proposed method has been tested on real car driving data, and the calculated position from the inertial navigation system in four maneuvers has been compared with the Extended Kalman Filter method outputs. According to the results of the experiments, the proposed algorithm has improved the position estimation by 60% on average during 30, 60, and 120 seconds without GNSS signals, compared to the inertial navigation system based on Extended Kalman filter.
    Keywords: Inertial navigation system, GNSS, LSTM, deep learning, Extended Kalman Filter, IMU
  • MohamadHossein Ashouri, Saeed Fakhte * Pages 55-63

    The mutual coupling issue between two Right Hand Circularly Polarized (RHCP) Magneto-electric (ME) dipole antennas is addressed in this study. To mitigate this issue, a Metasurface Polarization-Rotator (MPR) Wall is employed, resulting in effective minimization of the coupling effects. The innovative antenna design, with high gain, shows promise for 5G applications. It consists of two electric dipole plates with triangular corners positioned at the top, along with two plates acting as magnetic dipoles perpendicular to the ground plane. Additionally, the presence of four plates on the outer periphery of the antenna contributes to the improvement of the circular polarization (CP) performance of the antenna. The feeding structure is configured in a V-shape. Integration of the metasurface polarization-rotator led to a significant reduction in mutual coupling. On average, the mutual coupling is decreased by more than -20.5 dB, reaching impressive values of -45 dB at 2 GHz, -55 dB at 3.1 GHz, and -40 dB at 3.7 GHz when the MPR wall is placed between the ME antennas. The antenna demonstrates promising performance in terms of impedance bandwidth, with a remarkable value of 61.4% for |S11| < [-10dB]. Furthermore, the axial ratio bandwidth for AR < [3 dB] is 63.36%, representing an 11% increase compared to the configuration without the MPR Wall. The maximum right-hand circular polarization gain achieved by the antenna is 9.91 dB at a frequency of 3 GHz. Additionally, the maximum front-to-back ratio (FBR) is 37.6 dB at a frequency of 2.5 GHz. By comparing and analyzing the simulation results for the scenarios with and without the MPR Wall, it becomes evident that the MPR Wall does not significantly affect the parameters of gain, front-to-back ratio, and impedance bandwidth.

    Keywords: Mutual Coupling, magneto-electric dipole antenna, Wideband, High gain, Metasurface Polarization -Rotator
  • Maryam Ghelichkhani, Seied Ali Hosseini *, Seyyed Hossein Pishgar Komleh, Alireza Siadatan Pages 65-73
    In order to connect two binary and quaternary systems, it is necessary to use binary-to-quaternary (B2Q) and quaternary-to-binary (Q2B) converters. These converters convert numbers from logic 2 to 4 and vice versa.  In this paper, we designed a new binary-to-quaternary converter circuit using CNT transistors. In this circuit, the Power Delay Product (PDP) has been reduced to14.59% and 15.39% compared to best previous works (Ref [22] and Ref [23]). Also, this circuit has better driving ability and temperature stability than best previous works. The simulation results using Stanford's 32 nm CNTFET model in HSPICE software are at a voltage of 0.9 V.
    Keywords: CNTFET, multi-level circuit, B2Q Converter, mixed radix system
  • Maryam Amirabadi Farahani, Mohammad Haeri * Pages 75-89
    This article reports on a method for detecting disconnection between electric vehicle parking lots during charge management and uncertainties and how to deal with these issues. In this study, each parking lots has an aggregator that can exchange information with other parking lot aggregators through a communication graph. A cyber-attack or communication failure may cause a problem in the connection between the aggregators and their information exchange. To detect the loss of contact between the aggregators or uncertainties, a method based on the mean field game is developed through a distributed consensus algorithm. Since the number of vehicles in every parking lot, power consumption and generation are uncertain, the smoothness of the network load curve is disrupted. so, in this work an online optimization based on receding horizon concept is proposed to monitor network load every hour. However, due to the complexity of online calculations and disconnection detection, the optimization is implemented in an event-based manner. Although several distributed event-triggered methods have been introduced recently, these methods generally require state estimators to calculate the event-triggered error, the latest states and the threshold which increases the computation cost. However, the proposed event-triggered control method only requires mean field game information to compute the event-triggered conditions and requires less computations. To have convergent game, a time-varying network topology is suggested when the communication of parking lots is lost and the disconnection event is triggered. To validate the effectiveness of our method, we conduct computer simulations that demonstrate their achievements.
    Keywords: Aggregative games, Consensus Algorithm, electric vehicles, Event trigger, Switching topology
  • Mehdi Habibollahi, Ali Saghafinia * Pages 91-104
    The relationship between the amount of energy consumption and the circuit speed to change the design efficiency is an important challenge in designing digital circuits. Adders are essential components of computing circuits that play an important role in computing speed. This article proposed a new design for a single-bit current mode full adder using the field effect transistors based on carbon nanotubes to enhance the speed and reduce the occupied space on the chip. The correct combination of the majority function, the current mirror technique, and the sum value on carry reduced the delay of all adder circuits. The simulations have been done by HSPICE software and based on the provided standard model of 32 nm with CNTFET technology. The proposed design has improved by 55% in terms of delay. The PDP level in the proposed design has decreased by 63% compared to the previous designs.
    Keywords: Full adder, current mode, Field Effect Transistors, Carbon nanotubes, Majority function
  • Kirti Sharma, Pawan Tiwari *, Sanjay Sinha Pages 105-115
    Data represents a compendium of information that perpetually expands with each passing moment, contributed by individuals worldwide. Within the domain of medical science, this reservoir of data accumulates at an almost exponential rate, doubling in volume annually. The emergence of advanced machine learning tools and techniques, subsequent to a substantial evolution in data mining strategies, has bestowed the capacity to glean insights and discern concealed patterns from vast datasets, thus enabling extensive analytical pursuits. This study delves into the application of machine learning algorithms to enhance societal well-being by harnessing the transformative potential of machine learning advancements in the domain of blood glucose concentration estimation through regression analysis. The culmination of this investigation involves establishing a correlation between glucose concentration and hematocrit volume. The dataset employed for this research is sourced from clinically validated electrochemical glucose sensors (commonly referred to as glucose strips). It encompasses diverse levels of both glucose concentration and hematocrit volume, the latter being furnished by an undisclosed source to ensure copyright compliance. This dataset comprises four distinct variables, and the aim of this research involves training the dataset using regression techniques to predict two of these variables. Our results indicate that when utilizing linear regression, the R2 score for GC is approximately 0.916, whereas for HV, it reaches around 0.537. In contrast, employing the support vector regressor yielded R2 scores of about 0.961 for GC and 0.506 for HV.
    Keywords: Estimation, Correlation, analysis, Regression, Healthcare, enlightenment, Machine Learning, Quantum leap, Data mining, Insights
  • Davood Nazari Maryam Abadi, Ali Moarefianpour *, Nima Mahdian Dehkordi Pages 117-136
    In this paper, the problem of model-based finite-time bounded event-triggered control for distributed fuzzy T-S systems is presented. For this purpose, the whole network model is embedded locally in both the controller and the remote telemetry unit. In order to estimate the states of the plant between two consecutive events, a fuzzy observer has been used. Model-based state estimation reduces the state error and consequently leads to reduction of data transmission instants. By the network model and event triggering block which are placed locally in each remote telemetry unit, the time of data transmission on the distributed network is determined. Finally, the finite-time boundedness of the closed-loop system has been investigated using MATLAB software for a centralized and a distributed system, respectively.
    Keywords: Event-Triggered Control, Finite-Time |Bounded, Fuzzy T-S Control, Model-based control, uncertainties
  • Aqeel Breesam *, Ali Zalzala, Esraa Najjar Pages 137-143
    The rising use of digital imaging applications has recently boosted the demand for different image compressing algorithms. Picture compression is used to reduce unnecessary information from an image. We can store the vital information of a picture via image compression, reducing storage space, time, and transmission bandwidth. Although the results of lossy compression reconstruction are not similar to the original image the compression technique is essential to save a large amount of memory and increase the speed of transmission, especially when dealing with images. In this research a database of different five images were considered namely; woman, car, Lenna, peppers, and house with sizes of 33, 47, 40, 44, and 51 Kb respectively. The compression was fulfilled by Walsh transform with four compression ratios 5%, 10%, 15%, and 20%. The Walsh transform performed well and gave the highest average PSNR of 29.1904 during a 10% of compression ratio.
    Keywords: Walsh Hadamard transform (WHT), Image compression, Peak signal-to-noise ratio (PSNR), Mean square error (MSE), Database Image
  • Mohd Syahrin Amri Mohd Noh, Ghazali Omar *, Mohd Syafiq Mispan, Fuaida Harun, Zaleha Mustafa Pages 145-163
    Silicon wafers have been widely used in semiconductor manufacturing, and chipping issues often highlighted during wafer dicing which affects device performance and reliability. The phenomenon of chipping has been observed to have detrimental effects on die strength, leading to the potential of crack formation. Cracks became a major concern because its sometimes undetected during testing and had been reported to cause malfunctions at user applications. This study aims to comprehensively analyze the fragile behavior of silicon concerning its chipping and flexural strength performance, providing valuable insights for engineering applications. The research employed new wafer mounting techniques, including chipping analysis, a three-point bending test and scanning electron microscopy (SEM) to reduce silicon die chipping and increase the flexural strength by evaluating the novel semi and full sandwich wafer mounting techniques. The study demonstrated that the implementation of novel full sandwich mounting technique had improved significantly the silicon die chipping and flexural die performance among all the wafer mounting techniques.
    Keywords: chipping, silicon, Three-point bending test, wafer dicing, wafer mounting
  • Nayeem Ahmed *, Afrid Araf, Sheikh Rakeen Pages 165-178
    This paper focuses on the hybridization of two renewable energy sources (i.e., PV Panel and Bi-Cycle Dynamo) and the use of a Multi-Input SEPIC converter for highly efficient output from them. This paper has also presented the comparative assessment between the conventional Single-Input Single Output SEPIC (Single Ended Primary Inductor Converter) and Multi-Input SEPIC converter based on performance analysis. Both the Multi-Input SEPIC that is designed in this literature powered by the hybrid architecture of power sources and the Single-Input SEPIC are run in boost mode. The project is mainly developed with a view to facilitating the isolated rural islands and regions detached from the on-grid connections. The findings and assessments of this study are corroborated by the MATLAB/Simulink simulation results and optimally designed prototype built for miniature applications. The Multi-Input SEPIC topology has been developed in such a way that it is functional with an input voltage of 12.1V exactly and gives an output voltage of DC 53V approximately at the output terminal. To get the maximum voltage at the output from the designed circuit, the duty cycle of the converter recorded is almost 81.49%. The renewable energy sources (RES) that are used to build the prototype are Photovoltaic Panel and Bi-Cycle Dynamo. Due to the limitations of the PV Panels to generate power during the night and gloomy weather, the Bi-Cycle dynamo works as the backup power source. In performing the hardware and software analyses, various intermittent situations, solar irradiation, seasonal change, day-night phase, and other parameters are taken into consideration on the input terminal of the converter. An efficiency of 91.6% is obtained from the proposed hardware field-test analysis.
    Keywords: SEPIC Converter, Hybrid Renewable Energy Sources, Voltage Optimization, Rural Insular Area, Bi-cycle Dynamo, Solar Energy, Off-Grid Solution
  • Dahlan Abdullah *, Andik Bintoro, Cut Ita Erliana, Muhammad Fauzan Pages 179-185
    The flood detection tool uses the HC-SR04 sensor made specifically to detect the water level in the river. This tool is a series of simple tools that utilize the esp32 microcontroller and media monitoring in real time via a smartphone. The flood detection tool using the HC-SR04 sensor consists of a tool frame made of iron, the HC-SR04 sensor and a panel box. The main power source for this tool is a 12 V 5 Ah battery that comes from a 50wp solar panel. This tool is equipped with various electronic components including an LCD, HC-SR04 sensor, relay and stepdown module. The designed tool will be placed on the riverbank, the HC-SR04 sensor acts as a real time measure of water level. When the sensor has detected a predetermined water level, the sensor will send data to the microcontroller. The microcontroller will process the data and send it to the virtuino application which will display it virtually on the smartphone. Along with this, the data will also be displayed on the LCD
    Keywords: Flood Detection, Virtuino, HC-SR04, sensor, Microcontroller
  • Pardhu Thottempudi *, Vijay Kumar, Nagesh Deevi Pages 187-197
    This paper aims to provide a comprehensive examination of the Brain-Computer Interface and the more scientific discoveries that have resulted from it. The ultimate goal of this review is to provide extensive research in BCI systems while also focusing on artifact removal techniques or methods that have recently been used in BCI and important aspects of BCIs. In its pre-processing, artifact removal methodologies were critical. Furthermore, the review emphasizes the applicability, practical challenges, and outcomes associated with BCI advancements. This has the potential to accelerate future progress in this field. This critical evaluation examines the current state of BCI technology as well as recent advancements. It also identifies various BCI technology application areas. This detailed study shows that, while progress is being made, significant challenges remain for user advancement A comparison of EEG artifact removal methods in BCI was done, and their usefulness in real-world EEG-BCI applications was talked about. Some directions and suggestions for future research in this area were also made based on the results of the review and the existing artifact removal methods.
    Keywords: EEG, BCI, eCG, EMG, EOG
  • Gleb Vasilyev *, Oleg Kuzichkin, Dmitry Surzhik, Aleksandr Koskin Pages 199-204
    stationary operation of a thermoelectric system is often not economical due to the impossibility of soft adjustment of the microclimate. The aspect makes relevant the research of the transient modes of Peltier elements and thermoelectric systems built on their basis. The study of the dynamic properties of thermoelectric devices and systems with high accuracy requires the use of complex high-order models, which makes it difficult to obtain analytical solutions. Using the spectral approach and piecewise linear approximation, analytical formulae of the transient modes of Peltier thermoelectric elements for various deviations of control currents are obtained. The calculated relations of piecewise linear approximating functions, spectral densities and time forms of input and output parameters of the model under study are presented. Time diagrams of the temperature response of a particular Peltier element in linear and nonlinear
    Keywords: Peltier element, transition mode, piecewise linear function, Spectral method
  • Wayan Suparta *, Asmalia Ahmad, Asniza Abdul Tharim Pages 205-215
    This paper presents the engineering design of an unmanned aerial vehicle (UAV)/drone hexacopter and optimizes the PID (Proportional-Integral-Derivate) values for the Pixhawk 2.4.8 (PX4) flight controller. The design phase begins with component selection and identification, with the goal that the drone can carry loads up to 3 kg. Then install the main components and test the construction results. An analysis of the experimental results of the PID PX4 controller with no load and with load was performed. Results from direct field experiments with a home-built hexacopter show that the default PID must be tuned to be able to lift a load with a specific target.
    Keywords: Hexacopter Drone, Pixhawk 2.4.8, PID Controller Tuning, Delivery Packages
  • Ramadevi Kalluri, Prabha Selvaraj * Pages 217-240
    Plant phenotyping is one of the recent research areas that play an essential role to develop a better understanding of plant traits, genotypes, stresses, and other related features. It is regarded as essential concept as it facilitates development in several fields such as botany, agronomy, and genetics. Plant phenotype helps in acquiring relevant information about plant organs and whole features that allows the farmers to make informed plant cropping decisions. It includes the use of Deep Learning (DL) which is part of a machine learning technique that makes use of several processing layers to provide reliable outcomes from abstraction. DL-based approaches are highly useful in providing a sufficient amount of data related to plant strapping, stresses, and growth indices. Deep learning approaches are highly efficient in analysing plant phenotype and characterizing the phenotyping aspects by classifying the plant stress datasets into open, labelled, and broad-spectrum.  In this paper, a review work makes an attempt to explore the efficiency of deep learning and filtering approaches in plant phenotyping.  The recent works related to the DL principles have been utilized for digital image–based plant stress phenotyping. Then a comparative assessment of DL tools against other existing techniques, with respect to decision accuracy, data size requirement, and applicability in various scenarios. Therefore, it is strongly recommended in the study to use the imaging data process so that there is the attainment of accurate information from training datasets by using high-throughput systems like UAVs and other autonomous systems.
    Keywords: Plant Phenotyping, deep learning, Structural Phenotyping, Physiological Phenotyping, Temporal Phenotyping
  • Bao Tran Le Tran, Lanh Chu Van * Pages 241-251
    In this paper, benzene-core photonic crystal fibers (B-PCFs) with non-uniform air holes for the square lattice are studied by controlling the air-filling factor of the air holes in the first ring and increasing those up to a maximum for external holes. Commercial Mode Solutions software is used to study the influence of structural parameters on the nonlinear coefficient, dispersion, effective mode area, and attenuation. It can be seen that this design affects the linear and nonlinear properties at the same time. In other words, it has a crucial effect on either the near-zero flatness of dispersion or going up the nonlinearity and falling the loss of the B-PCFs. The above benefits make the optimized fibers suitable for supercontinuum (SC) generation applications.
    Keywords: Benzene-core photonic crystal fiber, Nonlinearity, Hollow-core, Square lattice, Dispersion, Optimize characteristic simultaneously
  • Nur Mostaman *, Erwan Sulaiman, Mahyuzie Jenal Pages 253-263
    The Hybrid Excitation Flux Switching Generator (HEFSG) has gained significant popularity in recent times owing to its relatively simple remarkably efficient topology. To optimize the performance of the generator, recent advancements and emerging patterns in mathematical modeling and software simulation, along with the utilization of optimization techniques, have facilitated the development of a novel methodology for electrical machine design. This study investigates the configuration and optimization of a Hybrid Excitation Flux Switching Generator, focusing on the rotor, armature coil, and field excitation. The optimization process involves multiple sequences for each component, employing the Local Optimization Method as an iterative approach to determine the optimal sequence that yields the highest output efficiency. Through the investigation of six rotor sequences, two armature coil sequences, and two field excitation coil sequences, a detailed optimization process was conducted. Consequently, the final output voltage of the HEFSG gains a 1.10% increment of voltage compared to the initial outcomes. Several sequences have influenced the output voltage performance of the generator during the optimization process. Therefore, modifications to the design of the arrangement contribute to the expansion of the operational range of the generator.
    Keywords: Optimization, HEFSG, Generator, FSG, Rotor, Armature coil, Field Excitation Coil, LOM
  • Prathap R *, Srilakshmi Uppalapati, Ganesh K Pages 265-282
    In terms of applications and research, VANET (Vehicular Ad-hoc Networks) communication is becoming more popular. Existing VANET communication protocols try to improve network performance but fail to consider security issues. Attackers exploit the vulnerabilities of VANET communication protocols. Providing security to the VANET is still a challenging task because of the vehicles' mobility and their short communication range. Based on our study, we found that wormhole attacks are the most common type of attack on VANET communication. The existing security solutions are inadequate to prevent or detect external wormhole attacks on VANET communication because these solutions do not consider important parameters such as vehicle mobility, neighboring ratio, and node mobility to detect the wormhole attack. To address external wormhole attacks in VANETs, we propose a two-level fast and efficient distance-based external wormhole detection and prevention system in this paper. In our proposed solution, we consider the vehicles' mobility, geographical location, and distance parameters to identify and isolate external wormhole attacker nodes. For effective monitoring of wormhole attacks, we use a dynamic threshold value for suspecting external wormhole attack links and then use the hop count metric to detect and prevent it. We used SUMO with NS2 simulators to compare our proposed system with existing works, and our simulations show that our proposed security solution outperforms existing works in terms of throughput, PDR, and jitter in a wormhole attack VANET environment.
    Keywords: Vanet, FEDEWPS, wormhole attack, vehicles mobility
  • Anitha Rani Palakayala *, Kuppusamy P Pages 283-309
    Parkinson's Disease (PD) is a neurological disorder that causes progressive loss of brain cells. Despite the fact that there is no known cure for this neurodegenerative disease at present, early diagnosis and treatment may improve the quality of life. Magnetic Resonance Imaging (MRI) detects structural changes related to dopamine deficiency in PD. To categorize MRI scans as Healthy Control (HC) or PD, this study proposes an ensemble of Deep Convolution Neural Network (DCNN) models. Initially, we have used DCNN models using augmentation and transfer learning, to classify as PD or HC. In the next stage, we applied a classifier fusion ensemble approach to enhance the overall result of the classification model. The proposed model is trained using the data collected from PPMI database, while assessed on custom dataset which is created using the data collected from Lalitha Super Specialty Hospital (LSSH). Finally, it was observed that the developed ensemble model produced an outstanding performance by plotting an overall accuracy of 99%, while the transfer learned EfficientNet B1 DCNN model stood in the second position, achieving a remarkable accuracy of 98%. This study serves as a significant step forward, providing valuable insights for researchers and clinicians engaged in the domain of clinical image assessment using deep learning techniques.
    Keywords: deep learning, Ensemble Learning, Neurodegeneration, Parkinson’s Disease, transfer learning
  • Lingappa Jaklair *, Guduri Yesuratnam, Pannala Sarma Pages 311-322
    This paper deals with the problems related to power stability issues in the isolated wind-solar based renewable energy system. The stability issues in such system are produced mainly due to unbalanced loading, distortion of load current and intermittent nature of renewable energy sources. To overcome these problems, A voltage source converter (VSC) is used in proposed literature. The VSC is driven using a time domain-based signal decomposition algorithm consisting a 3-phase dual reduced order generalized integrators frequency Locked loop (DROGI-FLL). The 3 phase DROGI FLL has the inherent abilities of noise rejection and error tracking, which improves the system stability and maintains the voltage and frequency of the system constant under transient conditions. The transient conditions are simulated by varying the loading condition, solar irradiance and wind speed in the system. Apart from stability issues this system also deals with the issues related to reactive composition, natural current compensation and operation of SPV system at maximum power point (MPP). A perturb and observe (P&O) maximum power point tracking (MPPT) is employed for the working of solar photovoltaic systems (SPV) system at MPP and to maximize the utilization of renewable energy. The neutral current compensation is achieved by using a star-double delta transformer.  This entire system is developed and tested in the MATLAB/SIMULINK environment.
    Keywords: voltage source converter, Power Quality, DROGI-FLL, Time domain-based signal decomposition, Wind-solar energy system, Load balance, Demand Response
  • Mohammed Ghribi *, Zine Eddine Ternifi, Ghalem Bachir, Michel Aillerie Pages 323-333
    The aim of this paper is to introduce a novel microinverter design that is based on the DC/DC converter, Buck. This structure is intended to provide a sinusoidal voltage to low-power autonomous loads embedded in vehicles, which require excellent waveform quality. In the first part, the paper presents the new topology, its modeling, and the control strategies employed. Two control strategies are discussed, including the classical control with a PI regulator and the sliding mode control with the step function. The dimensioning is carried out using the Ziegler-Nichols method, and the stability is ensured using the Lyapunov method. In the second part, the paper comments on the results obtained from different simulations in terms of harmonic distortion (THD), efficiency and control robustness with different sources, such as battery and photovoltaic panel.
    Keywords: Micro-inverter, buck, Discharge circuit, Lyapunov
  • Yousef Pendashteh, Seied Ali Hosseini * Pages 335-347
    Extremely efficient successor and predecessor circuits are suggested in this article using 4 CNTFETs. They have much less interconnections and complexity compared to the best previous circuits. The proposed circuits are designed by combining digital and analog techniques for the first time. They can be expanded for all MVLs like ternary, quaternary, pentaternary, and so on. The proposed designs for quaternary logic reduce the transistor count from 25 to 4 in comparison with the best previous works. Interestingly, in MVLs with more level logics, this difference will increase dramatically. This advantage leads to low complexity and costs. The accurate operation and great performance of introduced circuits are illustrated and their superiority is proved. Additionally, a quaternary half adder is founded on the presented successor and predecessor. The simulation results, which are acquired by comprehensive simulations utilizing Synopsys HSPICE and the 32 nm plenary CNTFET model of Stanford, show that proposed successor and predecessor circuits with only four transistors work accurately. According to these outcomes, in the proposed half-adder, not only the transistor count reduces 32%, but also it has 40% better PDP and 42.05% better EDP in comparison with the best previous work. Also it is more stable against process variation and robust in a wide range of temperature variation.
    Keywords: Multi-valued logic (MVL), carbon nanotube FET (CNTFET), nanotechnology, successor, predecessor circuits