Showing papers by "Umashankar Subramaniam published in 2022"

COVID-19 Detection Based on Lung Ct Scan Using Deep Learning Techniques

Abstract: SARS-CoV-2 is a novel virus, responsible for causing the COVID-19 pandemic that has emerged as a pandemic in recent years. Humans are becoming infected with the virus. In 2019, the city of Wuhan reported the first-ever incidence of COVID-19. COVID-19 infected people have symptoms that are related to pneumonia, and the virus affects the body's respiratory organs, making breathing difficult. A real-time reverse transcriptase-polymerase chain reaction (RT-PCR) kit is used to diagnose the disease. Due to a shortage of kits, suspected patients cannot be treated promptly, resulting in disease spread. To develop an alternative, radiologists looked at the changes in radiological imaging, like CT scans, that produce comprehensive pictures of the body of excellent quality. The suspected patient's computed tomography (CT) scan is used to distinguish between a healthy individual and a COVID-19 patient using deep learning algorithms. A lot of deep learning methods have been proposed for COVID-19. The proposed work utilizes CNN architectures like VGG16, DeseNet121, MobileNet, NASNet, Xception, and EfficientNet. The dataset contains 3873 total CT scan images with “COVID” and “Non-COVID.” The dataset is divided into train, test, and validation. Accuracies obtained for VGG16 are 97.68%, DenseNet121 is 97.53%, MobileNet is 96.38%, NASNet is 89.51%, Xception is 92.47%, and EfficientNet is 80.19%, respectively. From the obtained analysis, the results show that the VGG16 architecture gives better accuracy compared to other architectures.

Lithium-Ion Batteries—The Crux of Electric Vehicles with Opportunities and Challenges

Abstract: With the widespread use of lithium-ion batteries in a wide range of consumer electronics products, the CE industry has undergone a dramatic shift. The Li-ion battery has emerged as the heart of electric cars, and the focus has now shifted to the automotive sector. Liquid crystal displays have evolved over time to meet the demands of automobiles. International research groups and the performance of production electric vehicles are used to discuss and inform vehicle-driven battery targets. There is still a lot of room for improvement in terms of energy, life expectancy, cost, safety, and fast-charging capabilities for LIBs suited for the automotive sector. In this study, a review of lithium-ion battery applications in electric vehicles is presented.

Feasible Evaluation and Implementation of Shunt Active Filter for Harmonic Mitigation in Induction Heating System

Abstract: This paper proposes a shunt active filter (SAF) for harmonic mitigation and reactive power compensation in an induction heating (IH) system. The high-frequency switching in the resonant inverter of the IH system generates a considerable number of high-frequency harmonics. The latter flow back to the supply side and causes a wide variety of problems. The predominant ones are the deterioration of the power quality, distortion in the grid voltage, and malfunctioning of the protective equipment. These harmonics need to be attenuated as per the IEEE 519-1992 and IEEE 519-2014 standards. To overcome these problems, an SAF based on the modified version of instantaneous power theory was placed between the power supply and the IH equipment. Moreover, the proposed model could work in unbalanced and non-sinusoidal input voltage conditions, as well as dynamic conditions with changing reference currents. The feasibility of the proposed SAF-based IH system was verified by a series of simulation results and a comparative analysis of the THD of the input currents. The power quality issues were successfully addressed, which signifies the ability and effectiveness of the proposed model to mitigate the problems caused by harmonics and enhance the power factors.

Optimization of emission cost and economic analysis for microgrid by considering a metaheuristic algorithm‐assisted dispatch model

Abstract: Electric vehicles (EVs) have witnessed a steady and continuous rise in the last few years. The increased prevalence of EVs results in high demand for electricity from the power grid and it can be effectively handled by combining EV charging infrastructure with renewable energy sources (RES). However, the intermittent characteristic of RES introduces an additional challenge to the power grid. This paper proposes a metaheuristic algorithm called self‐adaptive elephant herd optimization algorithm (SA‐EHO) to achieve the desired dispatch model of microgrid (MG) with EVs and RES. The proposed algorithm optimizes both economic and emission cost of MG which comprises diesel engine (DE), solar photovoltaic (PV), EVs, fuel cell (FC), wind turbine (WT), and loads. The output constraints associated with a distributed power supply such as power limits of distributed generators (DG) and charging of EVs followed by its discharging are subjected to optimization. Finally, the performance of the proposed model is compared and proved over other existing models. Especially, a minimal total cost of the proposed model is 16.79%, 21.58%, 20%, 26.67%, and 1.33% superior to traditional HSA, GA, EHO, MSA, and FS‐MSA models at 100th iteration. Furthermore, the reserved cost of the proposed scheme for the third scenario is 77.78%, 90.91%, 85.71%, 90.91%, and 90.91% superior to existing models such as HSA, GA, EHO, MSA, and FS‐MSA, respectively.

Design of Novel HG-SIQBC-Fed Multilevel Inverter for Standalone Microgrid Applications

Abstract: The growth of distributed power generation using renewable energy sources has led to the development of new-generation power electronic converters. This is because DC–DC converters and inverters form the fundamental building blocks in numerous applications, which include renewable integrations, energy harvesting, and transportation. Additionally, they play a vital role in microgrid applications. The deployment of distributed energy resources (DERs) with renewable sources such as solar has paved the way for microgrid support systems, thus forming an efficient electric grid. To enhance the voltage of such sources and to integrate them into the grid, high-gain DC–DC converters and inverter circuits are required. In this paper, a novel single-switch high-gain converter (HG-SIQBC) with quadratic voltage gain and wide controllable range of load is proposed, the output of which is fed to a modified multilevel inverter for conversion of voltage. The overall performance of the newly designed converter and inverter is analyzed and compared with the existing topologies. A prototype of the investigated multilevel inverter is designed and tested in the laboratory. Development and testing of such novel topologies have become the need of the hour as the grid becomes smarter with increased penetration of distributed resources.

Solar PV network installation standards and cost estimation guidelines for smart cities

Abstract: For smart cities, the successful large-scale implementation of solar PV technology, Quality Certification and Standards are mandatory. The International Electrotechnical Commission (IEC) is a global organization for standardization consisting of all IEC national committees. The IEC PV standards comprise IEC technical committee 82 solar PV Energy System (IEC TC82) which develops and adopts all Photovoltaic related standards. There are nearly 80 standards applicable to photovoltaic and five working groups in IEC TC82. For necessary safety requirements ‘Quality and Standards’ technologically need to be revised and up to date. This paper presents PV standards developed by various technical committees worldwide, mainly focusing on various IEC PV standards, gaps identified by them and the recommendations provided by the committee in recent times. The breakup costs of the various sub-systems of a PV installation with an illustrative example for India is also discussed. It is intended to provide a guideline for consumers and investors, opting for solar PV installations.

Efficient Multi-Phase Converter for E-Mobility

Abstract: The recent growth of battery-powered applications has increased the need for high-efficiency step-up dc-dc converters. The step-up conversion is commonly used in several applications, such as electric vehicle (EV); plug-in hybrid electric vehicles (PHEV); photovoltaic (PV) systems; uninterruptible power supplies (UPS); and fuel cell systems. The input current is shared among inductors by paralleling the converters; resulting in high reliability and efficiency. In this paper; a detailed analysis for reducing power loss and improving efficiency is discussed. In continuous conduction mode; the converters are tested with a constant duty cycle of 50%. The multi phase interleaved boost converter (MPIBC) is controlled by interleaved switching techniques; which have the same switching frequency but phases are shifted. The efficiency of the six phase IBC model is 93.82% and 95.74% for an input voltage of 20 V and 200 V, respectively. The presented six phase MPIBC is validated by comparing it with the existing six phase IBC. The result shows that the presented converter is better than the existing converter. The prototype of the two phase and six phase IBC is fabricated to test the performance. It is found that the output power at the load end is highest for the 5 kHz switching frequency.

Design of Boosted Multilevel DC-DC Converter for Solar Photovoltaic System

Abstract: Integration of renewable energy sources to the grid-connected system has in fl uenced scholarly research in recent times to evolve solutions for power electronic conversion. Particularly, solar photovoltaic (SPV), being a resource available throughout the year, demands needful research to meet the demand for industrial applications. To facilitate SPV, multilevel inverters (MLIs) and cascaded H-bridge inverters (CHBIs) are proposed in the literature to meet the power requirement. However, these circuits su ff er from e ffi ciency loss, economic aspects of DC sources usage, and switching losses. Hence, in this research, a new power converter topology is projected to improve the overall e ffi ciency of SPV systems. Further, a three-level approach involving (i) SPV Panel-Temperature Reduction (SPV-PTR) Setup, (ii) Boost Multilevel Direct Current Link Converter (BMLDCLC), and (iii) use of e ff ective snubber modules (SM) are e ff ectively handled to promote the industry readiness of the proposed system. From a detailed system investigation, it is seen that the proposed arrangement has minimized the power loss to ensure better quality in output. Furthermore, the software-based results and hardware setup of the planned comprehensive converter have shown promising results in terms of (i) reduced voltage stress, (ii) reduced total harmonic distortion (THD) without fi lter component, and (iii) reduced power loss. It is observed that the experimental setup has reported a 12.9% of excess heat removal, 5% decrease in harmonics, and 33% switch reduction than the existing MLI schemes. In addition, the proposed setup is suggested to apply for industrial purposes indicate its e ffi cacy to be a solution in real time.

Moth-Flame Optimization for Early Prediction of Heart Diseases

Abstract: Heart disease is among the leading causes of mortality globally. Predicting cardiovascular disease is a major difficulty in clinical data analysis. AI has been demonstrated to be powerful in deciding and anticipating an enormous measure of information created by the health domain. We provide a unique method for finding essential traits employing machine learning approaches in this paper, which enhances the effectiveness of identifying heart diseases. Decision tree (DT), support vector machine (SVM), artificial neural network (ANN), and K-nearest neighbor (KNN) are the classification techniques used to create the proposed system. Ensemble stacking integrates the four classification models to create a single best-fit predictive model using logistic regression. Many explorations have been directed at the identification of cardiac infection; however, the exactness of the outcomes is poor. Accordingly, to further enhance the efficiency, Moth-Flame Optimization (MFO) algorithm is proposed. The feature selection strategies are used to improve the classification accuracy while shortening the execution time of the classification system. Medical data are used to assess the probability of heart disease based on BP, age, gender, chest ache, cholesterol, blood sugar, and other variables. Results revealed that the proposed system excelled other existing models, obtaining 99% accuracy in the Cleveland dataset.

Investigations on the Modulation Strategies for Performance Improvement of a Controlled Wind Energy System

Abstract: The challenges faced in an isolated wind energy conversion system (WECS) are larger transient times, high steady-state error, and larger harmonic content. To overcome these issues, an adaptive voltage controller (AVC) along with the load current observer (LCO) could be the better proposition. However, the AVC and LCO, in conjunction with the conventional space vector pulse width modulation (SVPWM) technique to operate the three-phase inverter of WECS, would not be able to further improve these parameters. This paper proposes the use of the unified voltage SVPWM (UVSVPWM) technique along with the AVC and LCO, which could improve the transient behavior by about 30% as well as reduce the harmonic content of the load voltage and current by about 70% and 2%, respectively. This paper considers an isolated WECS connected to the linear load, which is operated under balanced as well as unbalanced load conditions. The proposed control technique is verified for both the balanced and unbalanced cases using MATLAB/Simulink.

Reduction of Healthcare Resources by Classifying Patient's Data

Abstract: The number of electronic health records, or EHRs, being collected by medical care offices is steadily increasing. In dealing with patient information input that was traditionally documented in paper-based structures, it was customary for attendants to bear obligation. Precision is crucial when it comes to comprehending consideration and upgrading this massive amount of data helps the overall structure. Taking into account the cost of social insurance and the rate of interest. When asset criteria are inconsistent, the consideration conveyance approach is less effective. A sufficient number of patient groups will be required to consider the patient flow method utilizing a replica display or a logical model. They also mentioned that developing these groups based on the needs of patients' assets is becoming increasingly beneficial. Patients are placed together in emergency clinics based on the sort of medical care they received since they have similar assets after surgery. In any case, the loss of patients who are reliant on post-agent wards is very diverse and the average loss is the worst prediction of each entity's loss. Our goal is to lessen the vulnerability of patients' asset requirements and we do this by grouping patients into asset client groups that are comparable. Using electronic patient records, we construct a two-arrange characterization model to group patients into lower fluctuation asset customer bunches. Patients can be placed in lower changeability asset client gatherings using a variety of measurable indicators. However, because it has certain distinct advantages, arrangement and relapse tree (AART) evaluation are becoming a more popular technique for dissecting medical services data. Furthermore, we discovered that a subset of the variables, such as the key endorsed approach code, the confirmation point, and the working expert, might reveal up the level of 53:43% of the variation in patient duration-of-stay (loss). By lowering the vulnerability of patients' loss expectations, we may more effectively manage the quiet stream and achieve higher throughput.

Bearing Fault Detection For Water Pumping System Using Artificial Neural Network

Abstract: Bearing fault is a frequently occurring problem. In induction, motor and vibrational results can give data on the machine's health condition. The collection of vibrational signals through an external sensor is very costly. Conventional methods are time-consuming and expensive. Some of them are less efficient and require some energy to choose the signal. To overcome the problems machine learning (ML) algorithm-based model can determine the failure at an early stage. This work develops the Matlab Simulink model for the identification of bearing fault in induction motor-based pumping systems. In the experiment, the vibrational data have been collected by losing the bearing to some extent. The vibrational data have been applied in the Simulink model for verification of the system failure. Where ML-based artificial neural network (ANN) combined with genetic algorithm (GA), other ML algorithms like K-nearest neighbor (KNN), support vector machine (SVM), and decision tree is applied and compared based on evaluation parameters like precision, F1 score, and sensitivity to find out the best-suited algorithm. The result shows that the accuracy of ANN is more than other ML algorithms.

Energy Efficient Outdoor Lighting System Design: Case Study of IT Campus

Abstract: Outdoor lighting solutions are a key to safe urbanization. It is important that assessment of lighting is done at the design stage keeping in mind the crucial parameters like environment friendly design, light pollution, and energy efficiency. These crucial parameters are met by meticulously using lighting standards at each and every stage of lighting design. This process further naturally inclines in realizing sustainable lighting solutions for the future generations. In this paper, a case study of an Information Technology (IT) campus which requires suitable outdoor lighting system is considered. Outdoor lighting is very essential from the security perspective and also for the people and vehicle movement during night time. Hence it is necessary to design and implement an effective lighting scheme. To achieve this it is necessary to perform extensive lighting calculations and simulations in DIALUX software platform.

A Study Exploring Opportunities to Utilize Wind Charge in Bangladesh

Abstract: Wind charge is produced naturally in space. Knowing the nature of space charge is important for various purposes. If the charge is received for the whole year long by an antenna, then, in this process, the charge can be used as electric power. In the present work, the nature of wind charge in space in the cities of Dhaka and Narayanganj was observed. A net-type antenna was used to receive the wind charge. The results show that the power decreased with the antenna gaps. The output power increased with the antenna’s gross area and the speed of air. The overall output power was increased with humidity. It was shown that wind with a higher distance from the ground carries more electric charge. This research reveals that electric charge is inhomogeneous in wind, and it is collectable as electric power. The nature of the results is similar in both cities, but the output power is larger in Dhaka than in Narayanganj. This alternative power source is environmentally friendly, and it reduces thunder and lighting due to charge being taken from wind. Thus, this creates a more danger-free environment.

Impact of the flipped classroom approach in engineering education: A course analysis

Abstract: Abstract: The main purpose of the paper is to present a course analysis on the use of the flipped classroom. The objective is to explore the advantages and challenges for students and instructors. The impact of flipped classroom learning is analyzed with a set of students in the classroom. Student perceptions of lecture based learning and flipped classroom is compared to explore and promote active learning methodologies. The challenges of implementing a flipped classroom were identified and addressed using a variety of student-related activities. Mobile app learning methods, scale-up classroom and effective laboratory courses were implemented for a set of students. The significance of the flipped classroom model is noted. It was found that it promotes a deeper, broader perspective on learning, improves critical thinking capabilities and team work skills needed for a successful engineering career. Keywords: flipped classroom, scale-up classroom, active learning, learning outcomes

A non‐isolated bidirectional high gain integrated multiport converter for grid tied solar PV fed telecom load

Abstract: A multiport converter (MPC) with a non-isolated high gain bidirectional port is proposed for the grid integration of solar photovoltaic array (SPA) fed telecommunication load. The SPA along with a battery energy storage (BES) meets the power demand of the telecom DC load and the excess/deficit power is exchanged with AC grid. The MPC feeds the DC link of a voltage source converter for bidirectional operation with the AC grid. The small signal analysis of the converter shows that its operation is stable. The SPA, BES, and telecommunication load are rated for lower voltages, consecutively reducing the complexity with series-connected SPA. The proposed MPC possesses the merits of high voltage gain, reduced inductor size, and reduced number of components. Moreover, a power flow management algorithm is devised for the proposed converter that regulates the DC voltage at the telecom load and ensures smooth power flow control among various ports. The MPC is able to operate at various modes by controlling the ports independently. The converter performance during steady state and dynamic operating conditions under various modes are analyzed with detailed simulation studies. An experimental prototype is developed and test results are demonstrated to prove the viability of the designed converter.

Performance Improvement of Grid Interfaced Hybrid System using Distributed Power Flow Controller Optimization Techniques

Abstract: The main aim of this paper is to introduce a framework for the design and modelling of a photovoltaic (PV)-wind hybrid system and its control strategies. The purpose of these control techniques is to regulate continuous changes in the operational requirements of the hybrid system;currently, in power system networks, the distribution of energy plays a major role in maintaining power reliability in distribution systems. In this study, the proposed hybrid system was incorporated with a combined PV and wind energy system. Maximum power point tracking (MPPT) methods have been proposed to achieve maximum efficiency from the designed system. In addition, this study focused on improving the stability of the hybrid system. To improve the power quality and transient stability of the proposed system, we introduce a novel control strategy called the distributed power flow controller (DPFC) implementation with an optimization technique called the lion optimization algorithm(LOA)technique. This LOA control technique was developed for the first time in the application of a DPFC controller in a grid-connected system. The control technique was developed using signals from the system parameters, that is, voltage and current. To tune these parameters, this study used fuzzy logic and lion optimization techniques. The proposed system with controllers was tested in MATLAB/Simulink and the results were compared.