This paper introduces an innovative approach to enhance feature selection and enable dual fault diagnosis in analog circuits. By harnessing the combined power of fuzzy logic and neural networks, this study presents a robust framework for precisely pinpointing faults within the circuit. The operational integrity of any circuit hinges upon its inherent parameters, and the presence of defective components within the circuit distorts its performance, resulting in output deviations. The techniques elucidated in this research not only identify the specific faulty component but also quantify the extent of its deviation from the original parameters. A comparative analysis between the efficacy of fuzzy logic and neural networks in addressing this challenge is expounded upon. To demonstrate the practical application of these methodologies, a Sallen-key Bandpass filter is employed as the circuit under scrutiny (CUT). A comprehensive fault dictionary, constructed via the extraction of pertinent features from the CUT, serves as the cornerstone of this study. Notably, the fault table is meticulously generated, utilizing a step size of +/- 5% for parameter value perturbation.
The Infrastructure less mobile ad hoc network is known as MANET and all MANET environment networks are having Good Sensor Nodes (GSN) initially and they become Bad Sensor Nodes (GSN) due to internal and external attacks. Therefore, the MANET environment networks have both types of GSN and BSN. These BSN are categorized into Malicious Sensor Nodes (MSN) and Residual Sensor Nodes (RSN). The formation of these MSN and RSN may degrade the performance efficiency of the entire MANET environment network. Therefore, it is necessary to detect and mitigate these BSN from the network. In this work, GSN and BSN are classified using the proposed CNN structure. This proposed system consists of feature computations, feature optimization through Ant Colony Optimization (ACO) Algorithm and the optimized features are classified through the proposed CNN structure. The performance of the proposed MANET system is analyzed using precision, recall, True Negative Rate, Accuracy, Packet Delivery Ratio (PDR) and throughput.
Convolutional neural networks (CNNs) have been widely used in medical decision support systems to accurately predict and diagnose various diseases. Because of their ability to identify relationships and hidden patterns in healthcare data, CNNs have been extremely successful in developing health support systems. One of the most important and useful application systems is in the prediction of Heart Failure Diseases (HDFs) by observing cardiac anomalies. Fundamentally, CNNs have multiple hyper parameters and various specific architectures, which are costly and impose challenges in selecting the best value among possible hyper parameters. Furthermore, CNNs are sensitive to hyper parameter values, which have a significant impact on the efficiency and behavior of CNN architectures. Datasets from Electronic Health Records (EHRs) have recently been used to diagnose a variety of diseases, including heart failure. In this paper, we proposed the Deep convolutional neural network algorithm (DCNN), which is one of the deep learning algorithms that has been successfully used to solve computer vision problems. In our work, EfficientNet-B0 is a type of DCNN model that is used with a transfer learning approach to recognize diseases in Heart Failure images. To determine the effect of transfer learning with fine tuning, we assessed the performance of all EfficientNet-B0 variants on this imbalanced multiclass classification task using metrics such as Specificity, Recall, Accuracy, F1-measure, and Confusion Matrices. However, the accuracy and parameters of EHRs-based HFDs diagnosis are limited by the lack of an appropriate feature set. The experimental results show that EfficientNet-B0 achieves higher accuracy 98.45% with fewer parameters than the five classical DCNN models, demonstrating that the DCNN-EfficientNet-B0 model achieves more competitive results on HFDs identification.
An article that presents a simple printed antenna array that includes a split ring resonator and a stub patch is offered here. A microstrip feed line is used to provide the antenna with power. The thin copper with a thickness of 0.0035 millimeters was used to construct the hexagonal SRR and the rectangular stub patch. Using CST Software, the antenna array is developed and then manufactured on a substrate made of FR4. Tri-band operation at 19.46 GHz, 29.16 GHz, and 40.14 GHz is attainable with the antenna array that has been presented, which is equipped with a split-ring resonator. The use of parametric analysis allows for the determination of the optimal dimensions for the proposed antenna dimensions. Validation of the performance of the proposed antenna is accomplished via the use of simulated scenarios. A presentation is made on return loss, surface current density, gain, directivity, three-dimensional radiation pattern, E-plane radiation pattern, and H-plane radiation pattern. Because of its small size, steady radiation pattern, high gain, ability to be applied across three bands, and excellent impedance matching, this antenna is particularly well-suited for use in satellite and radar communication communications.
In this study, silica/Kaolinite/silver nanocomposites were synthesized according to experimental design results, using the central composite design (CCD) method. Samples were synthesized by impregnation on the polyester fabric, to get an in-situ approach to make a new performance of the polyester fabric to protect the human body from dangerous magnetic waves. Initially, magnetic saturation of the designed specimens was tested and its optimum values were measured with a Vibrating Sample Magnetometer (VSM) device. Mechanical properties including tensile strength, friction, abrasion, hydrophobicity (drop absorption), bending, thickness, and Crease Recovery Angle (CRA) of polyester fabrics impregnated with different amounts of nano-composite components were investigated using Response Surface Methodology (RSM) and PLS statistical methods which can help to show the effect of variables on each other. FESEM, EDX, and FTIR analyses were conducted for raw polyester-impregnated nanocomposites using an in-situ method under optimum conditions. The results confirm that the polyester fabric impregnated with three- component nanocomposite by varying concentrations of silica, Kaolinite, and silver, can significantly enhance the properties of saturation magnetic, strength, abrasion, friction, hydrophobicity, bending, thickness, air permeability, and CRA.
Lane and object detection is the major concern of an autonomous vehicle or driver assistance to mobilize continuously without making any traffic congestions and accidents. In a complex traffic scene countries like India facing many challenges to enabling the intelligent transport system in end-to-end customer connectivity. In this work the major district road (MDR) type is considered to identify the driveable space for the host vehicle. The proposed novel work is the combination of lane lines and object detection by LaneNet with sliding window and YOLOv5. Prior to the detection method, for computational complexity pre-processing methods, ROI and bird eye top down views are carried out. The object bottom corner coordinate points and lane boundary coordinate points on the reference line is considered to calculate the space on both sides of a front object of host vehicle parental lane. Finally, we used the real-time data and the most available CULane, BDD100K and TuSimple public dataset to perform simulation of a proposed work. LaneNet with sliding window for lane detection and pertained YOLOv5 model for an object detection and localization with an accuracy of 97% and 98% respectively. The simulation's outcomes demonstrate that the precision of the driving space identification results, 80% to 92% on various datasets.
Vector-borne diseases in India show growing patterns which strongly affect the population of the nation. The government faces a major obstacle in disease prevention efforts. Every year many people throughout India suffer from these illnesses. The physical differences between geographic regions and ways of life make it difficult for current strategies to control diseases during their initial development stages. The project focuses on creating advanced methods based on machine learning to diagnose diseases caused by vectors. The planned investigation targets dengue rather than other vector-borne diseases because it has emerged as one of the most dominant pathologies in contemporary years. A total of five stages make up the proposed methodology beginning with Data Transformation after which Preprocessing occurs followed by Feature Scaling and Normalization and finally Dataset Partitioning to allow Model Development and Prediction. The proposed model brings forth an ability to identify dengue fever development throughout its stages. The proposed solution stands out because it identifies dengue fever during early stages while determining the disease severity using patient clinical information. A test of the model used Support Vector Machine (SVM), Decision Tree and Gaussian Na & iuml;ve Bayes Classifier, Logistic Regression and Random Forest Classifier algorithms for evaluation. A biosensor was used for extensive testing and validation which enabled the suggested technique to produce a 97.5% accuracy rate. The Gaussian Na & iuml;ve Bayes classifier achieved 97.5% accuracy although it had a root mean square error value of zero.
Transportation is a major source of energy consumption and pollution in today's globe. Although electric vehicles appear to be viable solutions to these issues, their energy management systems are complex and need to be improved before they can be used widely. One of the major and most complicated concerns the globe has ever faced is reducing global warming gases produced by burning gasoline for transportation in vehicles. Electric vehicles, which are power-driven by an electric motor that runs on energy stored in a battery pack, were introduced in order to alleviate the environmental catastrophe caused by global warming. In this research, we have proposed the energy management system for electric vehicles (EMSEV) to stable the energy from the battery pack optimally. Moreover, a Fractional Order Adaptive Integral Hierarchical Sliding Mode Controller (FOAIHSM) has been designed for the smooth execution and energy management of EV in terms of output voltage regulation, reference generation, and smooth tracking of current. The proposed methodology incorporates total power inflow and state of charge of the power sources to satisfy load demands. Simulation results on MATLAB/Simulink have been used to verify the proposed controller's effectiveness. EMS based on Fractional Order Adaptive Integral Hierarchical Sliding Mode (FOAIHSM) Controller reaches 95% efficiency, resulting in smooth electric car performance of 94%. Experiments have been carried out more effectively to compare the results obtained with those of simulations. .
Microbial fuel cell (MFC) is one of the most important renewable sources for energy supply and reduction of environmental pollution, which has affected by various adversities due to operating conditions. In this paper, there are several serious issues related to the stable operation of a microbial fuel cell that have considered in the design of the controller, including: 1- Nonlinear terms that are of hard type; 2- Uncertainty of the model which is of parametric type and includes changes in temperature, environment and concentration; 3- Disturbances into the system which are of both matched and unmatched types; 4- And noise on the fuel cell output which has different origins. Also, the nonlinear model of MFC has considered for a more accurate description of system dynamics. By using of output feedback, adaptive, and sliding mode methods, and developing an approximation based on chebyshev neural network, a novel robust hybrid technique has proposed for controlling MFC output voltage and power. Using cheby shev neural network which has a simple structure with a suitable computational volume, the uncertainties, disturbances and hard nonlinear terms have approximated, and the optimal weights of the approximation have obtained by designing adaptive laws. Also, the robust part of the controller eliminates the effects of estimation error and noise. The Lyapunov's theory has used to ensure the stability of the closed-loop system. Furthermore, simulation in MATLAB environment and making comparison with the recent three robust methods in a strong scenario shows the efficiency of the proposed control method.
Load Frequency Control (LFC) is a critical aspect of power system control that ensures the balance between the generation and load demand. The demand for efficient LFC has increased because to the growing integration of dispersed power and renewable energy sources. Smart control techniques have emerged as a promising solution to enhance the performance of LFC in an interconnected power system. Despite the fact that various studies and approaches on load frequency control have been presented previously, no research concentrated on reviewing the approaches and limitations in the control techniques for load frequency control. Hence, this review paper presents a comprehensive overview of the recent developments in smart control techniques for LFC, focusing on Energy Storage Systems (ESS), conventional (DL) techniques, and Deep Reinforcement Learning (DRL). The advantages and limitations of each technique are discussed, and a comparison of their performance is presented. The review also highlights the future research directions and challenges in implementing smart control techniques for LFC in an interconnected power system and provides some suggestions for further improvements to be done in the future for better load frequency control in an interconnected power system.
In the present work dwarf palm leaf extract (DPLE) has been prepared to be used as a corrosion inhibitor on carbon steel in a saline medium. The inhibitory efficiency of DPLE on the corrosion rate of carbon XC70 steel (CS) in a 3.5% NaCl solution has been studied by means of weight loss measurements, potentiodynamic polarization curves, electrochemical impedance spectroscopy, SEM and AFM microscopies. The results showed that the corrosion inhibition rate of carbon XC70 steel in the NaCl solution increases with the concentration of DPLE, and reaches up to 90% at 2.0 center dot 10-4 g.L-1 as the optimum concentration of DPLE. The inhibiting performance against corrosion was attributed to the formation a DPLE barrier that reduces the contact area between the carbon XC70 steel and the corrosive solution. The EIS analysis revealed that the presence of DPLE was found to decrease the double layer capacitance, with an increasing charge transfer resistance. The morphological analysis showed that upon adding DPLE in saline solution, the surface morphology of the metal becomes smoother due to the formation a protective layer adsorbed on the metal surface. This study showed that dwarf palm leaf extract acts as an efficient and eco-friendly inhibitor on carbon steel in saline medium.
Even though it's still unclear how anticholinergic medications and dementia are related, dementia is one of the biggest global health issues. The current study's goal was to conduct a thorough review and meta-analysis of any potential predictive implications anticholinergic medications may have on dementia risk. Dementia has been linked to both low and high anticholinergic medication loading. Additionally, medications and the risk of dementia from anticholinergics were related. Among the anticholinergic drug groups, antiparkinsonian, urological, and antidepressant medications raised the risk for dementia. However, cardiovascular and gastrointestinal medications may have preventive effects. These results highlight the significance of anticholinergic medications as a potentially modifiable dementia risk factor and outline the most effective course of treatment. In this work, Work implemented AI Based algorithms of random forest and XG boost algorithm for predicting sleep disorder and Dementia with a help of sensor.
With the growing global population and the intensification of agricultural production, the pollution issues associated with farmland drainage have become increasingly severe. The excessive discharge of nutrients, particularly nitrogen and phosphorus, has emerged as a major cause of water eutrophication and environmental pollution. Traditional treatment methods have struggled to effectively remove these pollutants, highlighting the urgent need for efficient, economical, and sustainable water purification technologies. Electrochemical materials, known for their efficiency, controllability, and environmental friendliness, are gaining attention in environmental remediation, especially for the purification of farmland drainage water. However, current research has largely focused on laboratory conditions, lacking validation in large-scale practical applications and facing challenges in technology integration, cost control, and long-term stability. This paper investigates the application of electrochemical denitrification and dephosphorization technologies for farmland drainage water purification. It comprises two main parts: the development of electrochemical denitrification and dephosphorization technologies tailored for farmland drainage purification, and the experimental methods for testing the purification of farmland drainage water using electrochemical materials. This paper aims to systematically investigate the application of electrochemical nitrogen and phosphorus removal technologies for purifying farmland drainage water. By optimizing key parameters such as voltage, electrode spacing, and pH, the optimal operating conditions were determined and validated through experiments on actual farmland drainage. The study found that the electrochemical technology performed excellently in removing organic pollutants, achieving chemical oxygen demand (COD) and biochemical oxygen demand over five days (BOD5) removal rates of 92.68% and 96.96%, respectively. Meanwhile, the removal rates for total phosphorus and total nitrogen were 35.15% and 35.08%, respectively.
In the power distribution network, spread of power electronic devices and nonlinear loads has been exacerbate the power quality (PQ) difficulties. D-STATCOMs plays a key role to serve as an active power filter which are commonly used to address these issues. The performance of the distribution network is increased by incorporating the renewable energy sources (RES) such as PMSG- based wind energy conversion system (WECS) with DC capacitor across the D-STATCOM. During the PQ disturbances without controller the THD of source and load currents are 7.46% and 15.32%, by using the PI controller THD's are reduced to 3.02% and 4.01%. The proposed controllers reduced THD with ANN-PI source and load current THD are 2.99% and 3.62% and Fuzzy granular controller are 1.56% and 2.37%. Whereas on the other side, the DC-link voltage with PI controller introduces more fluctuations and reduced voltage level with settling time about 0.1secs, by using ANN-PI voltage fluctuations get reduced and maintain constant voltage, in addition to that in fuzzy granular the magnitude of dc link voltage increased by 10% and settling time got reduced about 0.05secs. The non- sinusoidal source current and load current are tranformed to sinusiodal by the ANN-PI and Fuzzy granular controller and also the power factor is improved nearly to unity. The proposed controllers are designed and tested by using MATLAB/Simulink.
A PV module subjected to decrease in the solar irradiance conditions leads to the reduction of power production. Shading of modules is one of the problems encountered in the system which makes the system less productive. The cause of shading may be of different factors like buildings, clouds etc. Hence, in a PV system, the placement of modules in a grid arrangement is a vital part, which influence greatly in the generation of power. If the position of modules is rightly configured, then the possibility of incidence of solar irradiance on the PV module can be enhanced leading to higher performance of the system. Therefore, the existing conventional series parallel (SP) and total cross tied (TCT) configuration are reinforced with a novel network topology namely extended cross diagonal view (ECDV) is proposed. The design is that the modules electrical connectivity is unaltered whereas the module position is altered. With the proposed topology, different kinds of shading patterns are imposed to carry out the performance analysis. The conventional SP and TCT configuration performance are compared with the proposed method which gave a significant rise in output power. An increase of 20.55% is observed by the proposed ECDV method for the crosswise shading pattern. From the analysis, the proposed method proves to be the most suitable and efficient method for generating maximum power under shading condition in non- square matrix photovoltaic (PV) system.
Multi-biometric authentication systems have become a viable way to improve authentication performance in the current digital era. Several multi-biometric authentication studies have been carried out and published in the literature. The difficulties of separating real biometric information from fraudulent attempts and integrating biometric and non-biometric authentication methods in a "Deep Convolutional Neural Network (KCP-DCNN)" that makes use of Kernel Correlation Padding are highlighted in this paper. An efficient multimodal Biometric Authentication (BA) system that integrates fingerprint, signature, and face modalities is presented in the study. To get ready for picture improvement, the input images are first pre-processed using the "Radial Basis Function-centric Pixel Replication Technique (RBF-PRT)". This procedure uses" Log Z-Score-centric Generative Adversarial Networks (LZS-GAN)" to apply blurring, augmentation, and noise reduction techniques to improve the visual quality of photographs. Following this, Dlib's 68-point facial landmark extraction is performed using the enlarged signature, fingerprint, and enhanced face photos. Using a generative adversarial network (GAN) that generates new images using log Z-scores as feature representations, a Chaincodecentric method is used for minutia extraction. This is then used in the" FDivergence AdaFactor-centric Snake Active Contour Model (FDAF-SACM)" for contour extraction. Key features are then retrieved using KCP-DCNN for efficient classification. The user is authenticated if the categorization output is accurate after the Quick Response (QR) code produced from the retrieved points has been confirmed. A user identification recognition accuracy of 98.181% is attained by the created model. In order to improve the "Multimodal Biometric" (MB) system's authentication rate, the suggested approach makes use of a biosensor.
