This research work proposes a hybrid ultra-capacitor-battery energy storage technology for electric cars. The Quasi Z-source inverters (qZSIs) buck/boost feature allows the Hybrid ESS(HESS) to be integrated into the traction-inverter-system (TIS). The switch can be activated for a quasi-Z-source network with Zero Current Switching (ZCS) process. To automatically turn off all free-wheeling diodes, the inductor currents in the quasi-Z-source network run in Boundary-Conduction-Mode (BCM) or Discontinuous-Conduction-Mode (DCM). It is possible to eliminate the battery converter and lower the rated potential for the battery part and Ultra-capacitors. In different operation modes, the stable power distribution theory is explained. On a short time, scale, a frequency diving frequency diving coordinated control technique is intended to maximize the parameters like battery current stress and dynamic- power regulation.
A blockchain is an ever-growing list of records that are linked to each other in a distributed network. These linked records called ledgers are immutable in nature providing resistance to change. Blockchain provides a secure way of processing the data in a distributed environment. It was widely involved in crypto currencies in the earlier days and however its application in bit coin motivated and inspired other applications to adapt its concepts. Its application in healthcare requires blockchain to be highly secure, provide a more trusted environment than the traditional blockchain, that is by design should be an enterprise level blockchain by restricting access to the public. Hyperledger Fabric caters to all these requirements in providing a secure and distributed environment for healthcare systems. In healthcare there are a lot of fields where Hyperledger Fabric can be adopted, but the focus here is given to management of patient's medical records. Traditionally the medical records are either stored centrally in a database that is accessible to only the hospitals owning it, this creates a several of problems for patients. The aim is to consider the records are handled, how the patient will interact in the real world and design a system using hyper ledger Fabric to tackle major problems using smart contract.
Obesity, a serious chronic disease, is on the rise as a result of how easily food can be brought to our door steps. People's need for food grew, and at the same time, their anxiety about their nutrition also grew. This study offers an image-based calorie estimation system that asks the user to upload an image of a food item in order to calculate the estimated number of calories in the image. It is a multitasking system that displays weekly information on a user's calorie consumption and the number of calories that must be ingested to prevent obesity-related illnesses like cancer, heart attack, etc. To recognize complex pictures, a collection of food images with 20 classes and 500 images are built in each class. This study has developed a six-layer Convolutional Neural Network (CNN) architecture for the purpose of extracting the traits and classifying the images. The proposed food identification trials had an accuracy of 78.7% during testing and 93.29% throughout training. By using software designed to accurately estimate food calories from still images, users and healthcare experts may be able to more rapidly detect dietary practices and food choices connected to health and health concerns. Calorie calculation has been done by using photographs, however, it is difficult, and there is presently no publicly available program that can conduct both food estimation using images and provide health information about the individual.
This project uses software to generate forensic facial art by obtaining information directly from the human brain via a BCI headband. We can quickly cut the time necessary to design the victim's face by automatically picking the pre drawn structure. The above suggested approach will not only sketch the victim's face, but it will also search the criminal database at random to see if the victim's face has previously been recorded. First, we use the Brain Computing Interface Band to get the EEG signal from the witness's brain. The EEG data is then processed in bit Brain to categorise it into each instruction, and the classified signal is then moved to the next phase to choose the face portion. This study includes the previously collected pre-drawn facial components and categorized the images by this point. The CNN algorithm is significantly more accurate in classifying the images, and the classified images are saved with the trail in BCI computing to select the image in an accurate way. A categorized image data collection is used to generate the processed EEG signal. to discover the face region that is equivalent to an EEG signal. Drawing software was used to choose the selected face portion, which was then placed at the fundamental facial structure. When the painting is 40% complete, the face structure is compared to an existing criminal database to check whether the facial structure matches any previous crimes. This initiative aids in the identification of criminals and the creation of forensic art in considerably less time than the traditional method.
Nowadays the use of inverters is increasing tremendously in many applications. Multilevel inverters give the accurate output waveform as a nearly sinusoidal waveform. This paper gives an overview of different types of multilevel inverters i.e., Diode-Clamped, Flywheel, and Cascaded H-Bridge inverters. The proposed topology deals with the REDUCED DEVICE COUNT 9-LEVEL INVERTER, its operation, switching sequence, and control technique used, and gives a review of output waveforms and THD. It has many advantages over the conventional 9-level inverter such as a lesser number of switches, low THD, high efficiency, and low price.
Road damage occurs when the function and structure of road are unable to service the traffic above it optimally. In general, the damage is caused by flaws in planning and implementation, uneven maintenance, poor drainage, and poor road user behaviour. It has a negative impact on driving comfort, road safety, and vehicle condition, and it may cause a number of accidents. To address this issue, this study presents a Region-based Convolutional Neural Network (R-CNN) for locating the dangerous path. This type of neural network can find essential information in both time series and picture data is the RCNN. As a result, it is extremely useful for image-associate tasks including image identification, object categorization, and design recognition. A RCNN uses linear algebra methods such as matrix multiplication to discover patterns inside an image. Find the photographs first and pre-process them, then extract the features and choose them from the feature set of previously damaged images. Finally, categorise the captured photos to obtain the optimum result. When compared to other current approaches, the suggested method is more accurate.
Since reversible computing can lower heat loss and power consumption, it is a more intriguing research area today. Processor power and speed are significantly impacted by adders. Reversible Toffoli gates are used in the suggested structure in place of conventional logic for the look-ahead logic. The suggested structure is given a hybrid variable latency extend, which minimises energy usage without dramatically slowing down speed. Reversible computing necessitates an equal amount of input and output lines. Backward deterministic system is a technique recycled for nanotechnology, inadequate capacity of VLSI design, Photonic computer, and nondeterministic computing. Indicated work shows how a carry look-ahead adder with a reversible Toffoli gate can decrease trash production, gate count, and quantum efficiency in comparison to the current design.
Advancements have been made in the field of face recognition technology. Controlling aperson's attendance in real time via facial recognition technology. Face recognition is the process of recognizing a person by their facial characteristics. Various computer vision algorithms, including those used for face detection, expression recognition, and video surveillance, can make use of a person's unique facial features. A face detection and recognition- based attendance monitoring system might very well rapidly and accurately locate and identify people in photographs or video footage. In addition to being laborious to maintain, the time-honored practice of physically ticking off attendees is inefficient. This research work presents the working of a cascade classifier built with machine learning to improve the face detection results. This has been done by comparing the face images in the current image to a database of previously trained faces. The acquired image contributions are searchedfor a previously registered face, and once found, the person's attendance is recorded automatically.
Nowadays the use of IoT in healthcare is a rapidly growing field that has the potential to greatly improve the quality of care and efficiency of healthcare organizations. By utilizing sensors and other devices, healthcare professionals can collect more data to better understand and predict the physical and mental health conditions of patients. This can assist with early detection of diseases and provide more accurate diagnosis and treatment. Wearable devices, remote channels, and other remote devices can be used to gather data on a patient's vital signs and physical activity levels, which can then be analyzed to identify patterns and trends. This can help healthcare professionals to identify potential health risks and to improve the overall management of patients' health. Additionally, IoT can also enable remote care for patients in remote or underserved areas, which can improve access to healthcare for these patients. However, there are also some challenges and limitations associated with the use of IoT in healthcare, such as ensuring the privacy and security of sensitive health data and the data's accuracy and reliability.
In electric motor drives, speed regulation plays an essential part in describing the overall performance of the system drive. To control the motor speed of an Induction motor (IM), an indirect vector control method is implemented in this paper. Scalar control is a simple and effective technique, but it responds slowly to transients and is unsatisfactory for regulating motors with dynamic behavior. The currents are controlled via the field-oriented control (FOC) approach, allowing for quick reactions. This approach meets the demands of dynamic drives, wherein quick response is required. The flux location is calculated indirectly in indirect control methods by rotor speed and slip calculation. The indirect control technique has grown in popularity due to the lack of rotor flux position sensors and the capacity to work at low speeds. A PI controller is utilized in the speed controller to control the motor torque by producing quadrature-axis current reference iq*. The motor's flux is controlled by direct axis current reference id*. The IM is operated by a current-controlled PWM inverter. The designed model is simulated using MATLAB and the results show an accurate speed response of the IM motor.
Health monitoring system in general is an innovation which is adopted worldwide especially in the last decade. The Patients suffering from permanent disables are much required to monitor them for their survival. Such patients are facing lot of issues because in no time their health is at risk. It is very difficult to predict them as there is a need of nursing all the time too. This paper presents a mobile phone linked health monitoring device programmed and controlled by Internet of Things (IoT). The important takeaway is to provide a handheld support for the healthcare professionals can monitor or to be notified when they are even outside the hospital environment to ensure the safety of the disable patients. The sensors collect the necessary information which is been sent to the IoT server and linked with the Internet module. The system consists of sensory devices, data acquisition system (DAQ), a controller (ESP32) and a software application. The body temperature, heat rate per minute, Heart ECG, oxygen level and Blood pressure are constantly monitored and stored as a report. The same report has been sent to the professionals, mobile phone through developed IoT application. Additionally, a text message is sent to the senior doctors' mobile phone if the sensor data exceeds the threshold value. Therefore, a mobile phone linked health monitoring system for disable patients constantly monitor and notify the concerned person on time and save precious life.
Mobile Wireless Sensor Network (MWSN) is an active network completely self-governing from a fixed structure. Also, MWSN has small channel bandwidth, great node movement, and inadequate battery energy. In WSN, flooding raises the reliability regarding successful communication of a packet with greater overhead. The flooding consumes the network's resources quickly, especially in terms of the lifetime of the node and energy, etc. Unicast routing creates an extra overhead and increases energy consumption because unicast routing mainly focuses on only minimum distance. But, multicasting is used when the desired destination is at maximum distance and increases the network's reliability in terms of throughput. This paper's objectives are to develop an efficient and reliable protocol using multicasting and unicasting to overcome the issue of higher overhead due to flooding. Multicast Routing Technique for Augmenting (MRTA) routing efficiency in MWSN is introduced. This method aims to enhance the link stability of the route for raising the route lifetime and diminishing the routing load. This method discovers the stable path from several ways to decline the overhead and delay between the sender and receiver. Simulation results demonstrate that this approach improves the network lifetime increases the throughput, and minimizes the routing load in the MWSN.
Wireless Sensor Network (WSN) suffers from different malware attacks. Several traditional approaches are proposed for detecting fault nodes in WSN. It is necessary to redistribute the nodes and detect the fault nodes while modifying the WSN parameters. Faults can occur in the nodes interrupting the continuous communication process of the nodes in the WSN. The main reason for a fault is also the induction of the energy drain in a node to the maximum possible level, failure of links in communicating nodes due to bandwidth constraints, and attacks induced by the malicious nodes. To solve this problems, Extended Finite State Machine based Fault Tolerance (EFSM_FT) in the WSN is used to detect the fault sensor nodes in the WSN. In this approach, called the finite state machine (FSM), state prompted is modified while a set of situations are true. Every node can be absolutely to be faulty or not establishing on the sensor node states. Using sensor energy, drop rate, forward sensor rate, sensor obtained rate, bandwidth, and transmission delay, this method finds the node with the fault. As a result, this approach accurately detects the faulty nodes in the WSN. The simulation outcomes illustrate the proposed method has a better fault detection ratio and minimizes the false negative ratio in the WSN.
Breast cancer is one of the terrible diseases among women worldwide. Better classification of breast cancer saves more human life and gives more confidence to extent their survivability. Deep Learning plays a vital role to solve real world complex problems in almost all domains including medical image processing. In this study, a novel DeepCNN model is proposed to classify Breast Cancer with better accuracy. Further, hyper-parameter optimization using Random Search is implemented to optimize the number of epochs, learning rate, and a dropout rate of the proposed DeepCNN model. Finally, the results are compared with various pre-trained models such as VGG19, Resnet50, Resnet101V2, InceptionResNetV2, InceptionV3, and Xception. Among all, the novel custom DeepCNN model produces higher accuracy of 99.18% using Random Search optimizer.
The act of separating a population or set of data points into a few groups or clusters so that data points in the same group are more like each other and distinct from data points in other groups is known as clustering. The purpose of this study is to categorize the respondents to identify groups with similar attitudes about science and technology and analyze their views. The difficulties of cluster analysis, determination of distance measure, number of clusters, and database structure have all been noted as possible issues with cluster analysis. To explore the respondents' grouping tendencies, several clustering approaches such as K-means, Hierarchical clustering, and so on are utilised The Hierarchical Clustering methodology itself may provide the analyst with the ideal number of clusters; human participation is not necessary. Dendrograms provide in clear imagery that is useful and simple to comprehend The centroids are computed by the K-means clustering method, which then iterates until it finds the ideal centroid It presumes that there are already known quantities of clusters. The flat clustering algorithm is another name for it. Since the data is binary, the clustering methods may be used to group the respondents. The clustering methods will be applied to the survey data by tracking the resultant decisions. Currently, all clustering algorithms have been used and it has been discovered that the data contains three or four clusters, each of which specifies a voting pattern that may be of interest.
Blockchain is basically a conveyed database that contains records and public record for all exchanges or computerized occasions that have been performed and shared among members. Every transaction in the public record is checked by agreement of most individuals in the system. When the data is placed, the data won't ever vanish. Block chain contains a conclusive and undeniable record of each and every transaction ever. The primary rationale of this of this proposed work is safely store and keeps up with the patient records in cloud database. Healthcare is a data serious space where a lot of data is created, dispersed, put away, and got to day to day. The Blockchain innovation is utilized to safeguard the healthcare data facilitated inside the cloud. The blockchain that contains the clinical data and Distributed computing will associate different healthcare suppliers. It permits healthcare supplier to get to the patient subtleties all the more safely from anyplace. It secures the data from attackers. The data is encrypted before moving to the cloud. The healthcare supplier needs to decrypt the data before downloading the information. The information is finally obtained by involving cryptography in the encryption stage and can be simply acquired by the client and server. In this proposed study, a healthcare record-based protection and security of healthcare-related information in the cloud has been performed on java platform.
Image encryption has been an appealing and exciting area for researchers recently. Many techniques attempt to enhance the security of images for storing in social media. Chaos theory is frequently used for image encryption due to its unpredictable nature. The classical chaotic based encryption methods are complex in structure and difficult to implement for providing high level securities. In order to solve these issues, anew method for encrypting images using Chaotic Maps (CMs) is proposed in this paper. Initially, a Logistic-Sine-Cosine (LSC)based CM is employed to generate various scrambling functions, such as Zigzag transform, Magic confusion, and Row confusion, and then these scrambling functions are used to modify pixel values to identify a linearity element. Then, these adjusted values are mixed with additional arbitrary sequence generated using the Logistic Cubic Cosine (LCC) based CM. Finally, the pre-encrypted images are combined together so that the generated randomness is dispersed uniformly throughout them. The combination of CM with scrambling functions not only improves security but also accelerates the speed of encryption.
Artificial Intelligence can quickly identify hazardous viral strains in humans. To detect COVID-19 symptoms, AI algorithms can be used to train to examine medical images like X-rays and CT scans. This can help healthcare providers to diagnose the disease more accurately and quickly. AI helps examine data on the spread of COVID-19 andmake predictions about how it will likely spread in the future. Machine learning algorithms known as Convolutional Neural Networks (CNN) are highly effective at evaluating images. As a result, CNN could assist in the early detection of COVID-19 by evaluating medical images like X-rays and CT scans to spot the disease's symptoms. This article's main aim is to provide brief information on some of the CNN models to detect and forecast COVID-19. The models were purely trained with Chest X-ray images of different categorized patients. The COVID-19 prediction models like ResNet50, VGG19, and MobileNet give accuracies of 98.50%, 97.68%, and 93.94%, respectively. On the other hand, forecasting also plays a vital role in reducing the pandemic because it helps us to analyze the risk and plan a solution to avoid it. The model is trained with some forecasting techniques like Prophet, LogisticRegression, and SEIRD model based on a text-based dataset that contains parameters such as the number of people infected per day recovered per day and many more for visualizing the trends in forecasting, which help in decision-making to analyze risks and plan solutions to prevent the further spread of the disease.
This research study shows how IoT technology can screen local meteorological conditions and share that information globally. Weather shifts cause extreme rainfall. A flood monitoring system uses NODEMCU ESP8266 to store and retrieve data and inform authorities of rising water levels using ultrasonic sensors and LEDs. Soil moisture affects crop growth. Its microprocessor and sensor improve soil moisture monitoring. An earthquake warning system can detect the slightest vibration before a major earthquake. Due to industry and autos, air quality is getting worse. Air quality and chemical content must be assessed using IoT since it has changed so much. Connected devices and enhanced sensor technology have transformed traditional environmental monitoring into a cutting-edge Smart Environment Monitoring System (SEMS). This paper evaluates SEM aids and research investigations, including air quality, weather, soil, and seismic monitoring systems. SEM applications segment the examination, with a deeper dig into each section's sensors. Discussion findings and analyzed research patterns form the basis for the in-depth analysis that follows the comprehensive review and suggests key SEMS research implications. The authors studied how IoT, machine learning, and other sensorbased advancements have made environmental monitoring smart.
The modern data centre (DC) is a perplexing amalgamation of various mechanical, electrical, and control frameworks. The increasing number of possible working configurations and nonlinear interdependencies make it difficult to comprehend. To improve Data Center performance and execution, the neural network scheme can benefit from the genuinely active data. Machine Learning (ML) is emerging as the most suitable method of demonstrating DC execution and improving productivity by utilizing the existing sensor data. The data centre infrastructure, the tasks to be completed, and the desired profits are framed as a mathematical programming model, which can then be overcome by using alternative methods of searching for smart task programming. This paper proposes a novel method for maximizing the load on cloud infrastructure. This article has utilized the machine learning techniques to enhance the load balancing. The new proposed algorithm is evaluated and then compared against existing algorithms for analyzing the make span time, time taken to execute tasks, overall load on the virtual machines and total cluster utilization.
