RSS 1.0Conference Papers
Recent Submissions
-
75th Annual Meeting of the Division of Fluid DynamicsNanosecond repetitively pulsed (NRP) discharges are a promising technique to enhance combustion efficiency and control. Numerical studies are essential to improve the understanding of complex plasma-combustion interaction. Limited by the prohibitive computational cost of fully coupled detailed plasma mechanism and combustion chemistry, a phenomenological model taken from literature is further developed to study the behavior of a laminar methane-air flame under NRP discharges. The phenomenological model focuses on two channels through which the electric energy is deposited: 1) the ultrafast heating and ultrafast dissociation of O2 coming from the relaxation of electronically excited N2; and 2) the slow gas heating coming from the relaxation of vibrational states of N2. The energy fraction deposited to these two channels is governed by the reduced electric field (E/N) which cannot be accurately predicted without resolving ion transport. Electric field is instead determined by solving the static poisson equation between two pin electrodes with three tested geometries. The predicted flame displacement under plasma qualitatively matches the experimental result. The roles played by chemical and thermal effects are strongly dependent on the E/N profile. Higher prediction of E/N magnitude at the preheating zone results in stronger dissociation effect and modifies the flame morphology more than what a lower E/N prediction concludes.
-
A phenomenological model for the impact of nanosecond repetitively pulsed discharges on a laminar methane-air flameNanosecond repetitively pulsed (NRP) discharges are a promising technique to enhance combustion efficiency and control. Numerical studies are essential to improve the understanding of complex plasma-combustion interaction. Limited by the prohibitive computational cost of fully coupled detailed plasma mechanism and combustion chemistry, a phenomenological model taken from literature is further developed to study the behavior of a laminar methane-air flame under NRP discharges. The phenomenological model focuses on two channels through which the electric energy is deposited: 1) the ultrafast heating and ultrafast dissociation of O2 coming from the relaxation of electronically excited N2; and 2) the slow gas heating coming from the relaxation of vibrational states of N2. The energy fraction deposited to these two channels is governed by the reduced electric field (E/N) which cannot be accurately predicted without resolving ion transport. Electric field is instead determined by solving the static poisson equation between two pin electrodes with three tested geometries. The predicted flame displacement under plasma qualitatively matches the experimental result. The roles played by chemical and thermal effects are strongly dependent on the E/N profile. Higher prediction of E/N magnitude at the preheating zone results in stronger dissociation effect and modifies the flame morphology more than what a lower E/N prediction concludes.
-
An Affordable EOG-based Application for Eye Dystonia EvaluationThis paper evaluates the risk of ocular dystonia-a condition marked by excessive blinking-using electrooculography. A commercial bioamplifier is employed to capture the electrical activity of the eyes using dispensed surface electrodes. The continuous wavelet transform of the electrooculogram was estimated to identify the features related to involuntary eye-blinking behavior and make the classification. The signal processing is integrated into a novel application with a simple graphical user interface oriented to be used by physicians. The performance is evaluated using multiple evaluation measures. Results show that the proposed method succeeded in identifying an abnormal frequency of blinks with respective accuracy, precision, sensitivity, and specificity scores of 98.46%,96.51%,99.13%, and 96.41%.
-
EEG Signal based Schizophrenia Recognition by using VMD Rose Spiral Curve Butterfly Optimization and Machine LearningSchizophrenia is a mental illness that can negatively impact a patient's mental abilities, emotional propensities, and the standard of their private and social lives. Processing EEG data has evolved into a useful tool for tracking and identifying psychological brain states. In this framework, this paper focus on developing an automated approach for recognizing schizophrenia using non-invasive EEG signals. The EEG signals are segmented and onward decomposed by using the Variational Mode Decomposition (VMD). Each mode is termed a variational mode function (VMF). Onward, features from each intended VMF are mined based on a Rose Spiral Curve (RSC). The mined features are concatenated to present an instance. Afterward, the most pertinent features are selected using the Butterfly Optimization Algorithm (BOA). The selected feature set is conveyed to the classification module. Two classification approaches are applied in this study namely, the k-nearest neighbor (k-NN) and Random Forest (RF). The applicability is tested by using a publicly available EEG schizophrenia dataset. The highest accuracy of 89.0 % is secured for the case of RF.
-
Effect of MPPT Technique in Solar System Efficiency: A Case StudyWorldwide energy demand is growing fast because of the population explosion. Technological advancement paved the way toward utilizing renewable energy sources instead of fossil fuels as they cause harmful effects on the environment. Among all renewable sources, solar energy is a promising source. Solar energy is captured by photovoltaic (PV) arrays that convert the sunlight into electricity that powers the load. A vital component in the PV system is the DC-DC converter. Maximum Power Point Tracking (MPPT) techniques can be applied to the DC-DC converter to deliver maximum power from the PV array. Thus, the efficiency of the PV system increases. This paper compares the output of a PV system with and without applying MPPT techniques. The PV system installed at Effat university library's rooftop is considered a case study in this paper. Matlab/Simulink software is used for simulation. The results show that applying MPPT leads to a more significant PV power. The output PV power when applying MPPT has an efficiency of 99.93%, and when there is no MPPT, the output power has an efficiency of 21.21%. Also, the power, voltage, and current waveforms in the system with MPPT achieved less harmonic distortion than the system without MPPT
-
Design of Fuzzy Controller for a Hybrid Active/Passive Cooling System in Smart Homes with a WindcatcherGlobally, most building energy consumption is associated with heating, ventilation, and air conditioning systems (HVAC). Building energy consumption increased from 115 EJ in 2010 to nearly 135 EJ in 2021, accounting for 30% of global final energy consumption. In 2021, electricity will account for approximately 35% of building energy use, up from 30% in 2010. Space cooling, in particular, saw the greatest increase in demand across all building end uses in 2021, increasing by more than 6.5% over 2020. This study aims to set design guidelines to reduce energy consumption in building sector by proposing a hybrid active/passive cooling smart system. This can reduce energy consumed by the electricity grid by achieving natural ventilation through wind catchers. The later is a historical architectural element used in buildings to provide cross ventilation and passive cooling. The architectural modeling of the proposed system's design is conducted using Autodesk Revit. The smart controlling system is implemented with Fuzzy logic in MATLAB Simulink. Moreover, the accuracy of the system is improved by a PID tuning based on Backpropagation Neural Network. The results confirmed the effectiveness of the methodology used.
-
Design and Optimization of PID Controller based on Metaheuristic algorithms for Hybrid RobotsMetaheuristics optimization techniques are significant to search methods that are used to solve challenging Artificial intelligence (AI) problems. In hybrid robot control systems, Meta-heuristic optimization methods are widely applied. The major goal of this paper is to develop optimized PID control parameters to improve the performance of the hybrid robot control system. For that purpose, two optimization techniques followed by fine-tuning are proposed and simulated to get the optimized PID parameters. The first proposed optimization method applies the Satin Bowerbird (SB) optimization technique to optimize the PID parameters. The Crow Search Optimization (CSO) technique is applied to the SB results to improve the algorithm's performance and the PID parameters. The second proposed method applies the Emperor Penguin Optimization (EPO) technique for the optimization of the PID parameters. The results of both methods are fine-tuned. Moreover, a Kalman filter is used to improve the outcomes after and before tuning the PID parameters. Simulation results show that the proposed first method is more effective for the optimization methods of the PID controller, and its results outperform the results given by previously published research.
-
Design of a DC/DC Converter with a PID Controller and Backpropagation Neural Network for Electric VehiclesCurrently, global warming has become a major problem. The pollution caused because of conventional internal combustion engines are increasing dramatically. Electric Vehicles are good alternatives to conventional IC engine vehicles in promoting a green environment. Controllers, converters, and modulation schemes are needed to provide a safe and reliable power transmission from energy storage systems to the electric motor in electric cars. In this paper, a design of DC/DC boost converter based on a PID controller is proposed. Moreover, a Back Propagation Neural Network (BPNN) technique is applied to generate the optimal PID parameters before using the PID. The proposed DC/DC boost converter is simulated using MATLAB software. The simulation results proved that the proposed DC/DC converter with PID-BPNN achieved higher performance and a stable output voltage.
-
Efficient Tradeoff between Throughput and Energy Efficiency of Massive-MIMO Technique for Satellite Communication applicationsThe rapidly increasing demand for mobile communication over satellite platforms and its applications necessitates a significant effort on the part of researchers to fulfill the prospective requirements of wireless network infrastructure. It is predicted that traffic will increase by multiples of hundreds soon. Therefore, the network's capacity has to multiply with high energy efficiency (EE), which can be achieved using massive multiple-input, multiple-output (M-MIMO). An adaptive scheme that maximizes energy efficiency is proposed in this paper at maximum spectral efficiency. Also, an efficient tradeoff between energy efficiency and throughput is mainly proposed. The analytical and simulation results prove that the proposed multi-cell minimum mean square error (M-MMSE) precoding scheme provides the maximum EE and efficient throughput of next-generation networks and satellite communication utilizing M- MIMO. Hence, it gives the optimum and most efficient tradeoff between EE and the throughput of the M-MIMO system.
-
A proposed Array of Quadrifilar Helix Antenna for CubeSat applicationsRecently, small satellites such as Cube Satellites (CubeSats) have played an important role in modern communication and applications like imaging, remote sensing, education, wide-area measurements, deep-space research, and deep-space observations and communications. CubeSat antenna has primary functions in telemetry and command, communication, navigation, and inter-satellite links (ISL) operations. This paper investigates the radiation characteristics of proposed arrays of quadrifilar helix antenna as an application on a CubeSat in L-band. Firstly, the problem is formulated using the moment method for a four-element array on top of a CubeSat is considered. Then, the radiation patterns, array gain, and axial ratio are depicted. Finally, the characteristics of a cavity-backed four-element quadrifilar helical antenna on top of a CubeSat are investigated. Simulation results show that the antenna gain increases using the cavity while the axial ratio slightly decreases.
-
A Novel Web-Based Multi-Class Heart Disease Prediction Using Machine Learning AlgorithmsIn the present scenario, heart disease impacts human life very inadequately and raises the cause of death in the world. In order to prevent heart failure, an early precise and on-time diagnosis is very significant. Through the conventional medical record, heart disease diagnosis has not been considered reliable in many aspects. In this regard, the authors developed a novel medical diagnosis system using machine learning (ML) algorithms. The optimal set of features is considered to enrich the proposed heart disease multi-class classification superiority by utilizing the different feature selection methods. The authors utilized Random Forest (RF), Support Vector Classifier (SVC), K-Nearest Neighbors (KNN), Extreme Gradient (XGBoost), and ensemble learning classifiers (Stacking and Voting) to perform the multi-class classification. To determine and investigate the findings realized by ML algorithms, the performance measures such as receiver optimistic curve (ROC), accuracy, precision, f-score, and recall are considered and observed. In addition, the FLASK framework was implemented and deployed for the application programmable interface (API) and web page. The results reveal that the Stacking ensemble classifier accomplished exceptional accuracy and the ROC score of 93.5% and 0.99%, respectively. The suggested medical diagnosis system will assist doctors/hospitals in predicting heart disease risk aspects easily.
-
Power Quality Disturbances Classification Based on the Machine Learning AlgorithmsThis paper presents an approach for classification of the power quality disturbances (PQDs). The classification of real-time power quality disturbances (PQDs) is proposed in this work. The PQD signals are modelled based on the IEEE 1159–2019 standard. The outcome of the used PQD model is employed for analyzing the performance of suggested classification method. Firstly, the PQD signals are segmented and then each segment is further processed by machine learning based classifiers for identification of PQDs. The study is conducted for six major classes of the PQDs. The highest identification precision is secured by the Support Vector Machine classifier. It respectively attains the Accuracy = 94.32%, Precision = 84.55%, Recall = 84.33%, Specificity = 96.52%, F-measure = 84.19%, Kappa = 92.59%, and Area Under the Curve (AUC) = 97.83%.
-
Rechargeable Battery State Estimation Based on Adaptive-Rate Processing and Machine Learninghe generalization of the use of electronic systems and their integration in industrial systems and different aspects of modern life (internet of things, electric vehicles, robotics, smart grids), give rise to new challenges related to the storage and optimized management of energy. Lithium-on batteries perfectly meet this objective due to their good qualities such as high energy density, small installation size, low self-discharge and high supply capacity. However, their wide application requires further research on battery failure prediction and health management. Intelligent “battery management systems” (BMSs) employ the real-time estimation and control algorithms to improve the battery safety while enhancing its performance. Nevertheless, BMS are complex and require increased processing power which could lead to more power consumption. In this context, the present article provides a new approach for efficient prediction of the “Lithiumion” (Li-ion) battery cells capacities by analysing and exploiting the battery parameters, acquired by an event-driven module. It acquires the intended cells voltages, currents and temperature values during the charge-discharge cycles. The solution is based on the machine learning algorithms and event-based segmentation. The “National Aeronautics and Space Administration” (NASA) has provided a high-power Li-Ion cells dataset for the purpose of research and innovation. This dataset is used to test and evaluate the suggested approach. The evaluation of the overall performance of the chain has shown encouraging results of the proposed approach.
-
Hybridization of Wavelet Decomposition and Machine Learning for Brain Waves based Emotion RecognitionEmotion recognition has sparked the interest of researchers from a variety of disciplines. Studies have demonstrated that brain signals may be utilized to characterize a wide range of emotional states. Electroencephalogram (EEG) measures the cerebral activity. Therefore, by exploiting the EEG signals the emotion states can be determined. In this study the EEG signals undergoes through filtering, segmentation, Wavelet Packet Decomposition (WPD), feature mining, and classification. The machine learning algorithms used for classifications are “Decision Tree” (DT), “Support Vector Machine” (SVM), and K-Nearest Neighbor” (K-NN) algorithms are used for categorization. Their performance is compared for automatically identifying the emotion state. It is determined that the best performer is SVM. It has attained 98.2% accuracy, 97.3% precision, 97.3% recall, 98.7% specificity, 97.3% F1, 97.3% kappa, and 99.3% AUC.
-
Features Mining and Machine Learning for Home Appliance Identification by Processing Smart meter DataThe energy sector is changing as a result of digitalization and IoT advancements. The Internet of Energy (IoE) is developing to link many smart grid components and shareholders effectively. The use of smart meters is becoming more popular in this context. The automatic identification of appliances is one of the most important applications of smart meter data. Enumerated billing and dynamic load management are possible outcomes. This process is complicated due to the usage of many brands and types of equipment. For the purpose of automatically identifying significant home appliances based on their usage patterns, this study presents a novel hybridization of segmentation, time-domain feature extraction, and machine learning algorithms. While automatically categorizing six key household appliances of various manufacturers, the developed technique achieves 96.2 percent accuracy, 97.7 percent specificity, and 98 percent AUC values.
-
Epileptic Seizure Detection Using the EEG Signal Empirical Mode Decomposition and Machine LearningEpileptic seizures affect millions of people worldwide. Medical treatments exist to help lessen the severity of the damage caused by these seizures. However, people with epilepsy still struggle with unexpected seizures. People who experience epileptic seizures have Electroencephalogram (EEG) signals that show different features in comparison to a healthy brain. In this study, EEG signals are studied to detect the seizures. The incoming signals are denoised by using linear phase filters. In next step these are divided in fix-length segments. Then, each segment is broken down using the Empirical Mode Decomposition (EMD) into Intrinsic Mode Functions (IMFs). For an automatic identification of EEG signals, features are extracted from the collected IMFs and then processed using machine learning techniques. A dataset on epilepsy that is available to the public is used for evaluation. To determine which is the best predictor for the under-consideration dataset, four different classification methods are performed and the results are examined. The system achieves a classification accuracy of 96.70%.
-
Heart Disease Identification Based on Butterfly Optimization and Machine LearningThis paper aims to make use of The Physionet Challenge 2016 collection of normal and abnormal heart sound recordings that were classified by automated identification of PCG sounds to help detect heart diseases earlier and prevent incidents. People with heart diseases have been increasing and most of them lead to fatalities so the detection of sound signals through PCGs can be applied in machine learning models by extracting features from data. In this study, the dataset's recordings are segmented to be used in variational mode decomposition. Once they are decomposed, those means will be fused together into a set of features given to the Butterfly Optimization Algorithm which will conduct a selection of features. As the features are selected, MATLAB was used to test various machine learning algorithms. Results from Support Vector Machines (SVM) and artificial neural networks were used in this investigation (ANN). The ANN model, which had an accuracy rate of 94.8%, was the most accurate of them.
-
Non-Invasive BCI by using EMD and Machine Learning: A Metaverse Interaction PerspectivePeople with disabilities struggle to perform specific tasks throughout their daily life. However, BCI systems are developed to assist people struggling with motor impairment by transforming their thoughts into action. Non-invasive BCI systems use electroencephalogram (EEG) to record brain activities. In this study, we segment the EEG signals and then break the segment down into a few intrinsic mode functions using oscillation mode decomposition. Then the intrinsic mode functions are mined for feature extraction. The features mined are processed by different machine learning algorithms for categorization. Among the different algorithms, K-NN yielded the best results with an overall average accuracy score of 95.48%. This approach can be used in future to develop the brain driven metaverse interactive solutions.
-
Discrimination between benign & malignant breast tumors using a region-based measure of edge profile acutanceWe propose an adaptive measure of edge profile acutance to distinguish between benign and malignant breast tumors on mammographic images. The measure is computed using differences along normals to the boundary of the tumor region. Classification rates of up to about 95% were obtained in two-group and four-group classification (the latter using additional shape factors) of 39 tumors from the MIAS database and 54 tumors in an augmented database.
-
An Approach for Detecting Missed Tissue Proteins in Autoimmune DiseasesAutoimmune disease is a pathologic condition resulting from an induced error in the immune system leading to an autoimmune response with organ dysfunction or tissue damage. The discovery of autoantibodies in the blood is essential in the diagnosis of these diseases. Notice that the antibodies may not be the essential reason for the disease. It should be remarked that auto-antibodies are commonly found in all immunologically competent people and can increase during disease, infection, or injury. In some cases, autoantibodies can be the result, not the reason, of the disorder process. The existence of autoantibody responses has major significance in the diagnosis and prognosis of several autoimmune disorders. The goal of this work is to detect the set of missed tissue proteins that can be used in the diagnosis and treatment of a specific autoimmune disease. A hypothetical EDAS is generated. Ten thousand records are randomly created based on the mathematical model. The developed algorithm for missed tissue protein discovery is described. The presented tool can be used to diagnose autoimmune diseases in clinical laboratories.
