Today's complex civilization requires a steady and constant supply of electrical energy. It's crucial because electrical power creation necessitates effective management. This is because the amount of energy generated must equal the amount of energy used, also accurate and timely forecasting is required. In order to characterize future load in sufficient quantitative clarity, the estimated load demand is determined using previous knowledge. The load time series has a significant level of uncertainty, making it difficult to create an accurate short-term load estimate. Therefore, this work examines numerous load forecasting problems, as well as related recent constraints, utilizing various methodologies.
Our food security is built on the foundation of soil. Farmers would be unable to feed us with fiber, food, and fuel if the soils were not healthy. Accurately predicting the type of soil helps in planning the usage of the soil and thus increasing productivity. This research employs state-of- the-art Visual Transformers and also compares performance with different models such as SVM, Alexnet, Resnet, and CNN. Furthermore, this study also focuses on differentiating different Visual Transformers architectures. For the classification of soil type, the dataset consists of 4 different types of soil samples such as alluvial, red, black, and clay. The Visual Transformer model outperforms other models in terms of both test and train accuracies by attaining 98.13% on training and 93.62% while testing. The performance of the Visual Transformer exceeds the performance of other models by at least 2%. Hence, the novel Visual Transformers can be used for Computer Vision tasks including Soil Classification.
Nonlinear dynamics are critical in nuclear power reactors such as Advanced Heavy Water Reactor (AHWR). The core of AHWR must be governed and controlled by a suitable and reliable controller to deliver the demand electricity properly. In this paper, a fractional-order nonlinear controller (FO-NPD) comprises of fractional-order (FO), nonlinear (N), and proportional plus derivative (PD) terms are proposed and incorporated into the normalized point kinetics (NPKs) model of AHWR core for trajectory tracking performance, disturbance rejection, and noise suppression analysis. The FO- NPD controller gains are modified at run-time using the cuckoo search algorithm (CSA) based on a performance metric index termed as the sum of the integral of square error (ISE). The tuned gains get modified in real-time which makes the recommended controller more resilient and robust. To validate the obtained simulation results, a comparative performance study is carried out between the proportional plus derivative (PD), the fractional-order proportional plus derivative (FO- PD), and the nonlinear proportional plus derivative (NPD) with the proposed FO-NPD controller. And it is observed that the proposed controller demonstrates the superiority over other three controllers.
The present study employs a fair bidding blueprint acceptable in the Electricity market in real-time. For market players, reliable electricity prices prediction is crucial. Price projections are used by market participants to determine their trading strategies, portfolio allocation and risk mitigation. For proper incorporation, the appropriate bidding patterns have been analyzed so as to employ maximum revenue generation by the power producing companies. Contribution in terms of mathematical framework and modelling under certain constraints like seasonality, weather, previous loads and electrical prices, has been presented here, so that overall profits in the intraday market get amplified. A proper balance in the bidding aspects of Day-Ahead and Intraday Markets with incorporation of Deep-Learning in the electricity price data has been achieved, that too specifically using the concept of Artificial Neural Network modelling via MATLAB.
With the increased penetration of distributed generators (DGs) in the distribution system, the protection schemes based on standard inverse time characteristics are unsuitable as DGs alter the power flow from unidirectional to bidirectional and change the fault current level depending upon the operating mode of the microgrid. All of these result in the need for directional elements and communication links for proper coordination, which increase the cost and complication of the protection scheme. In this paper, the study of optimal relay coordination in a DG-based system is performed with user-defined characteristics, which is a logarithmic function of fault voltage and has a grading of relays independent of operating modes of the microgrid. Further, the performance of the voltage-based relay in terms of optimal settings and total operating time is compared to the conventional overcurrent relay. The optimal solution to the relay coordination problem is determined by a genetic algorithm (GA) and simulated annealing (SA), and the results of both the methods were compared for the user-defined and the standard relay characteristics.
Internet of Things (IoT) devices generate a lot of data periodically in the internet era. In order to process that data in real-time, high storage capacity and computational power is required to overcome the limitation of the low processing capabilitiesand storage ability of these devices. Cloud computing solution offers large storage capacity and powerful computational facility but increases the response time due to network latency. To address the issue, the idea of Fog computing was introduced by bringing the computational services closer to the peripheral device of the network. The internet-enabled e-healthcaresystem allows doctors to conduct remote patient monitoring and demands real-time decision-making for critical data. Smart decision-making for classifying the data gathered into high and low-risk data has been considered in some of the existing research. In this context, we designed a framework to analyzethe accuracy of e-healthcare systems by employing machine learning approaches for data classification in fog computing. Also, comparative analysis has been done, highlighting the performance of data classification approaches. Simulation resultsshow that K-Nearest Neighbor and Support Vector Machine classifiers perform better than other classifiers observed and fog performance is significantly high in comparison with cloud.
Nowadays modern society is over flooded with humongous masses of visual data. There exist many image analysis methods to dive into this sea of visual information. The constituents of these images and videos can be analyzed and further processed to recognize the useful information. The detection, identification, and localization of different objects could prove to be of mammoth use and can play a significant part in modern devices and technologies. This paper represents a comparative study of several entity recognizing methods like YOLO, Faster R-CNN and R-CNN over different parameters such as mAP, FPS, etc. This paper also introduces an intelligent system (robot) that is capable of localizing an object and following it in real-time. The required input image is provided by the ESP32 cam module which can be mounted on the robot. Machine Learning Algorithms are used for object detection. Position coordinates received are then used to locate, track and follow the moving object. Furthermore, it is of interest as it can scale down human tasks and help mortals to be aware of minute details about certain objects.
The simplification of higher-order interval system is implemented by applying the Cauer-second form and Mihailov criterion with the Kharitonov theorem. If the system is stable, the suggested method produces a stable reduced-order model. An example is solved to show the effectiveness of the technique. The responses show a close approximation of a higher-order system with the suggested reduction method. Further, comparing performance index and time domain specifications with recent methods validate the efficiency of the proposed approach.
This work investigates the issue of guaranteed cost control (GCC) based anti-windup design for the continuous-time systems with actuator saturation (AS). A quadratic Lyapunov function is used to establish the stability of closed-loop systems. The existence of dynamic output-feedback controllers is demonstrated using linear matrix inequalities (LMIs). The anti-windup compensator is designed to mitigate the detrimental “windup” effects of the AS. The proposed approach guarantees the asymptotic stability of the systems as well as provides an upper bound for the closed-loop performance index. A design problem of optimal anti-windup compensator gain with guaranteed cost is converted as a convex optimization problem. The usefulness of the proposed strategy is demonstrated with the help of an appropriate example.
The vagueness of wind speed, which disturbs the balancing environment of the electrical system, creates wind farm integration with the standing central electricity grid exceedingly problematic. In today's competitive power market, maintaining the quality of given electricity to customers, as well as maximizing profit for supplied power resulting from the integration of such sporadic renewable energy sources is extremely difficult. This study outlines an effective optimization strategy for maximizing consumer welfare and profit, as well as the profit of generation farms. Actual and predicted wind speed data has been collected in real-time. This study has been performed with modified IEEE 14 bus test system to authenticate and evaluate the efficacy of the presented approach using MATPOWER Software. The effect of wind farm integration on Locational Marginal Pricing (LP) has also displayed in the presented work.
The demand of wireless power transmission is increasing day by day, and resonant dual active bridge (RDAB) has been found to be useful because it helps in achieving high efficiency due to less conduction losses. The main advantage of the RDAB is that it needs less reactive power due to resonance. The resonant dual active bridges give us the advantage of bidirectional power flow, i.e., power can be transferred from source to load side or load to source side. The necessary reactive power needed for converter operation is supplied from the capacitor in primary side of a transformer. To show wireless power transfer instead of using linear transformer, loosely coupled inductors are used in this paper. The RDAB Capacitor-inductor-inductor-capacitor (CLLC) topology helps in achieving zero voltage switching for wide range of loads and it also results in less switching losses and reliable operation. In this paper two different topologies are compared through the MATLAB simulations and then using coupled inductors, bidirectional wireless power flow is shown through CLLC converter.
Viscosity measurement has wide ranging applications from oil industry to pharmaceutical industry. However, measuring viscosity in real-time is not a facile process. This paper provides an elaborate mathematical model and study of viscosity measurement in real-time using pressure sensors. For a given flowrate, a change in liquid viscosity gives rise to a change in pressure difference across a particular section of the pipe. Hence, by recording the pressure change, viscosity can be calculated dynamically. A mathematical model as well as a finite element analysis model has been presented to determine viscosity from flowrate and pressure difference. A set of pressure sensors can be placed at a fixed distance from each other to get the realtime pressure change, while flowrate can be obtained using a flowmeter. For the finite element analysis, the pressure sensors were placed 60 mm away from each other. The radius of the pipe was 19 mm. A mixture of water and glycerol, with different ratios, was used to provide variable viscosity.
In this paper, a PV integrated battery storage system (BSS) is implemented to maintain the power demand. The power demand is fulfilled by maintaining the output DC voltage constant by charging and discharging the battery accordingly. In the initial stage, a photovoltaic (PV) system along with Boost converter having Perturb and Observe (P&O) control has been implemented. In the next step, an isolated DC-DC Dual Active Bridge (DAB) bidirectional converter (BDC) is used to charge the battery in one mode and use the battery as a power source to power the load in another mode. In the later stage, battery charging and discharging has been discussed and it has been efficiently implemented. For various load applications required constant output voltage is being properly maintained. All the works have been successfully simulated in MATLAB-SIMULINK and it can be considered for future work as a reference for zero current switching (ZCS) and zero voltage switching (ZVS) switching of the bidirectional converter for a more efficient and economical way of using it for long term application.
Commercial usage of sea wave energy has been thwarted by the lengthy and labor-intensive technique necessary to harvest it. An innovative wave energy gathering method has been devised during this research. This idea employs a novel topology to eliminate the limitations of typical linear generators and to increase the continuous power supply, together with a simple design of floaters and a rotatory generator. An investigation into a cutting-edge technological system with the potential to reduce the costs associated with building and operating an ocean wave energy (OWE) power plant that is connected to the grid, while simultaneously improving operational efficiency and expanding business opportunities was the primary objective of this project. Installing and maintaining floaters along a coastline is simple because of their ability to be attached to the beach or any other man-made structure. A synchronous machine is used to produce electricity at this site, traditionally linear generators are used to extract energy from the ocean. With an average wave height of 0.8 meters, the sea still has the ability to generate and transport electricity with a high voltage and current value. Programming language Python and MATLAB software are used to analyze the project's output, respectively. Besides electrical generators, the emphasis here is on hydraulic motors.
A non-isolated multiport sott-switched bidirectional (BTPC) dc-dc converter for fuel cell electric vehicle applications is presents in this paper. The implemented design increases the efficiency of the traditional three-port converter (TPC) and widens its applications by including a soft-switching circuit, as well as a bidirectional power direction that allows the power storage device to be charged from the output. In addition, there are no soft-switching issues with any bidirectional three port converter (BTPC) transitions in any working condition. In the implemented architecture, a single inductor is employed in all power directions. Overall the conversion occurs in a single stage in all modes of operation and as a result the conduction losses get decreased. Coupled inductors are employed to control the soft-switching circuit, and tune the magnetic core and converter percentage. The various modes of operations of the proposed BTPC converter are emphasized in detail.
The electricity prices find wide applications in the present electricity markets. Generation companies use the forecasted electricity prices to plan their expenses, which helps the aggregator provide better consumer services. The market players also use it to strategize selling electricity to the distribution companies in the energy exchange market. The forecasted electricity prices are also used to implement Demand Response (DR) programs. DR programs reduce peak demand, which can be achieved by scheduling the loads. All these applications require accurate forecasting of electricity prices, but its volatile nature makes it challenging to make accurate predictions. Electricity price forecasting on hourly basis is presented using the Long Short Term Memory (LSTM) Neural Network model. LSTM model can extract highly complex relationships between parameters. It uses feedback property to predict electricity prices accurately. The proposed model predicts the prices based on the historical hourly prices, historical load demand data, and other quantities on which the prices depend. The accuracy of the model is compared with other baseline models. The model comes out to be the most accurate of all and has the lowest Mean Absolute Percentage Error (MAPE) and the highest R2 value.
The renewable energy sources (RES) have a major problem of reliability due to their intermittent nature and inability to provide continuous power. For this reason some backup mode of energy storage system is required, which needs to be bidirectional in order to absorb the surplus power and should supply the power when in need, either in short-term or long-term. The future electricity grid is predicted to have more capacity share of renewable energy sources. This poses the concern of grid stability and reliability. As a result of the integration of more renewable energy sources such as the solar PV into the grid, the fault tolerance capability of the grid will become a major concern. When a large solar PV will be disconnected from the grid due to a fault, it may lead to grid instability, economic losses and eventually grid failure. To reduce the imbalance in the distribution system especially during the fault, this paper proposes a solution in the form of grid scale battery storage which enhances the capability of fault tolerance by absorbing the solar PV active power and injecting reactive power. All the concerned simulations have been performed using the software MATLAB/SIMULINK.
In this work, a Fractional Order Proportional Integral Derivative (FOPID) controller is designed for the performance enhancement of a Pressurized Heavy Water Reactor working under step-back condition. The Reduced Order Model (ROM) for the PHWR system is obtained by the Optimal Hankel Norm Approximation (OHNA) method. The unknown controller parameters are tuned using the Nelder-Mead optimisation algorithm. A performance analysis is carried out to show the effectiveness of the designed FOPID controller.
Wireless charging technique using Inductive Power Transfer (IPT) are attractive for electric vehicles (EVs) since they are free from any contacts. The IPT systems of various power ranges from few watts to several kilowatts are used in transport applications like- electric bicycles, electric cars, electric trains and healthcare. Wide variations in load, leakage inductance due to airgap and switching frequency are major challenges in development of wireless charging system using the IPT technique. A series - series compensation technique based on Inductive Power Transfer is proposed in this paper to provide misalignment independent and constant charging at various vehicle load conditions, meanwhile reducing the complexity in control strategies and making the resonant circuit independent of misalignments. Using this proposed resonant circuit topology the vehicle charging also becomes bifurcation tolerant, hence making it a suitable charging topology even in the dynamic charging modes. Wide variations in load causes the variations in input currents for which the proposed compensation circuits are used, which can reduce these inrush currents. This paper also aims at the design of compensation circuit at resonance to reduce the inrush of current and making it efficient charging system using IPT (Inductive Power Transfer).
Sign language recognition can help hearing and speech impaired persons to communicate with the rest of the world. Hand gesture recognition is widely used in the entertainment sector for cars, gaming, and other devices. In the healthcare industry, clinicians can utilise this recognition technique to handle digital images instead of touch screens or computer keyboards during medical procedures where sterilization is necessary, as well as to automatically recognise surgical gestures. For real-time, image-based hand gesture identification, recent improvements in machine learning and object detection approaches provide improved and more efficient results. We used the deep learning-based object detection models YOLOX and YOLOv5 to work on five different hand gestures for recognition in our research project. Both models were released at the same time, but YOLOX was significantly more precise and faster. For YOLOv5, we got the highest mAP score of 98%, the lowest mAP score of 89.6%, the precision score of 98.8%, and the recall score of 82.6%. Our best mAP for YOLOX was 99.55%, while our last mAP was 98.5%. In this research study, we evaluated the performance of the two models and investigated changes in the architecture that enabled YOLOX to perform so much better than YOLOv5.
