Journal of Science Engineering Technology and Management Science

The rapid proliferation of electric vehicles has necessitated the development of resilient and intelligent charging infrastructures capable of accommodating fluctuating demand whilst preserving grid integrity. In this regard, hybrid energy storage architectures, particularly those combining battery energy storage with fuel-cell systems, have emerged as a promising solution for grid-interactive charging stations. This paper presents a decentralised power coordination framework for an electric vehicle charging facility integrated with photovoltaic generation, battery storage, and a proton exchange membrane fuel cell. A modified line-resistance-compensated droop mechanism is employed to regulate power sharing among distributed sources, thereby ensuring stable direct-current bus operation under both grid-connected and autonomous conditions. Particular emphasis is placed upon state-of-charge-oriented battery regulation to restrain excessive depth of discharge, while the fuel cell is constrained to operate within its linear efficiency region to enhance durability and operational economy. The proposed strategy facilitates unity power factor interaction with the utility grid, sustains acceptable harmonic performance during islanding events, and accommodates dynamic charging scenarios without reliance on centralised communication. The effectiveness of the control methodology is substantiated through comprehensive simulation and experimental investigations, demonstrating improved power-sharing accuracy, prolonged storage life, and uninterrupted charging operation under diverse operating contingencies.

Keywords : Electric vehicle charging station; Decentralised power management; Hybrid energy storage; Fuel-cell systems; Battery state of charge; DC microgrids; Droop control strategy

Author : Mr. J. KANTHAIAH,Mr. K. RAMAKRISHNA PRASAD,Chimmana Naga Sanjay,Esam Rakesh

Title : A DECENTRALISED HYBRID FUEL-CELL AND BATTERY ASSISTED POWER COORDINATION SCHEME FOR GRID-INTERACTIVE ELECTRIC VEHICLE CHARGING FACILITIES

Volume/Issue : 2026;03(02)

Page No : 1-9

This project focuses on automatic video generation using artificial intelligence to simplify the process of video creation. Traditional video production requires significant time, cost, and technical expertise, which makes it difficult for students and beginners. The proposed system converts natural language text into complete videos by generating scripts, voice narration using Text-to-Speech, captions, and relevant background visuals. Natural Language Processing is used to analyze the input text, while video processing techniques combine all components into a final video. The system works through a user-friendly web interface where users only need to enter text, without requiring any video editing skills. It reduces manual effort, saves time, minimizes cost, and is useful for students, educators, and content creators, demonstrating how AI can simplify multimedia content creation.

Keywords : Artificial Intelligence, Text-to-Video Generation, Natural Language Processing, Text-to-Speech, Automated Video Creation, Multimedia Content Systems

Author : GOWTHAM G,ANAS MOHAMMED M,SARAVANAVEL V,SATHISHKUMAR J

Title : AN ADVANCED ARTIFICIAL INTELLIGENCE FRAMEWORK FOR GENERATING VIDEOS FROM NATURAL LANGUAGE TEXT

Volume/Issue : 2026;03(02)

Page No : 10-21

The quick development of smart home technologies and the growing interest in sustainable energy solutions, have pushed to merge renewable micro-generation systems on IoT energy management platforms. This paper shows the smart house energy management system design and its implementation based on a free turbine turbulence in micro system. The device captures air flow and turns it into energy - a task that the IoT-based microcontrollers and sensors monitor, control and manage. Data on turbine output, household consumption and environmental parameters are streamed using lightweight communication protocols to a cloud hosted platform. It offers a dynamic load balancing control between renewable generation, grid supply and optional energy storage for optimizing household energy consumption. A homeowners-dashboard platform delivers energy production and consumption data; a breakdown of savings and enables the ability to remotely turn on and off appliances. The system also includes predictive analytics to improve efficiency by predicting demand and distributing energy as needed. The IoT-based smart energy management system proves the viability of the micro-level usage of renewable generation and intelligent control with scalability and modern living sustainability concept. Employing a turbo wheel micro-turbine demonstrates the abilities of small renewable power systems in minimizing grid reliance and creating sustainable lifestyles.

Keywords : Smart Home, Energy Management System, Turbo Wheel Turbine, Renewable Energy, Microcontroller (Arduino).

Author : V.MALLIKARJUNAREDDY, Mr.K.SAIMAN, A.V.BHARATHREDDY, CH. SASI, M.MOHITSAI

Title : SMART HOME ENERGY MANAGEMENT SYSTEMS USING A TURBO WHEEL MICRO TURBINE

Volume/Issue : 2026;03(02)

Page No : 22-28

Brain tumors are a serious health problem, and detecting them early is very important for successful treatment. Normally, doctors confirm tumors through surgery and biopsy, which are invasive methods. In this project, artificial intelligence (AI) is used to provide a non-invasive way of diagnosis. By using MRI brain scans, deep learning models are developed to classify gliomas, meningiomas, pituitary tumors, and healthy brain tissues. The main model used is a Convolutional Neural Network (CNN), which learns patterns from a large dataset of MRI images. To improve accuracy, preprocessing techniques such as normalization, scaling, and data augmentation are applied. These steps make the system stronger and reduce false positives and false negatives. The proposed model shows high accuracy in separating tumor and non-tumor cases, helping doctors make faster and more reliable decisions. This project demonstrates how AI can support medical professionals, reduce manual effort, and provide precise forecasts. Overall, it highlights the importance of automated systems in medical imaging, making diagnosis more efficient, trustworthy, and useful for early detection of brain tumors.

Keywords : Brain Tumor Identification , Deep Learning with CNNs , Healthcare Image Processing , Intelligent Diagnosis Systems , AI Applications in Healthcare

Author : SK.FAZARUNNISA BEGUM, Mrs. K. RAJYAM , Y. AMULYA , CH. HARSHA VARDHINI, G. ANUSHA, T. AMULYA

Title : AUTOMATED BRAIN TUMOR DETECTION USING IMAGE PROCESSING TECHNIQUE

Volume/Issue : 2026;03(02)

Page No : 29-36

The requirements for portability have required rapid development of low power electronic communication products. The variable gain amplifier (VGA) is one of the weakest components of any modern receiver in terms of a correct signal baseband processing. Most CMOS VGAs have low bandwidth and limited gain range. In this paper, two stage class AB VGA consisting of direct transconductance amplifier and linear transimpedance amplifier is designed in Silterra 0.13-µm CMOS technology using Mentor Graphics environment. According to the post-layout simulation findings, the VGA design achieves the highest gain range of -33 to 32 dB and the broadest bandwidth of >200 MHz. The VGA merely uses 2 milliwatts of power. a single DC 1.2 V source. The core chip area of the VGA is likewise only 0.026 mm2 which is also the lowest compared to current researches. For all contemporary communication gadgets, such a VGA will be a very helpful module.

Keywords : CMOS, Receiver, RF, VGA

Author : G. VENKATA BHAVANI PRASAD, Mr. P. SIVA KRISHNA, K. GOPI KRISHNA, CH.VENKATA SRINIVASA RAO, S. VENKATA SAI CHARAN TEJA, D. PAVAN KUMAR

Title : DESIGN OF A LOW-POWER COMPACT CMOS VARIABLE GAIN AMPLIFIER FOR MODERN RF RECEIVERS

Volume/Issue : 2026;03(02)

Page No : 37-45

Imaging methods like Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET), and Single-Photon Emission Computed Tomography (SPECT) have given medical professionals information about the soft tissues, structural features, and other aspects of the human body. distinct sensors collect distinct image information of the same part, and different imaging techniques maintain different properties. Better contrast, fusion quality, and perceived experience are the goals of the fusion. The outcome of the fusion should satisfy the following requirements: (a) the fused image should fully preserve the information of the source images; (b) the fused image should not generate any artificial information, such as artifacts; and (c) poor states, such as noise and misregistration, should be avoided. The suggested network is one of three CNN models used to compare patch similarity. Because its two weight branches are the same, the source image's feature extraction and activity level measuring methodologies are also same. This has some advantages over pseudo and 2- channel models, and the siamese model is also used in fusion applications due to its ease of training. After getting the weight map, the Gaussian pyramid decomposition is used, followed by the pyramid transform for multiscale decomposition, bringing the fusion process closer to human visual perception. Additionally, the decomposed coefficients are adaptively adjusted using the localized similarity-based fusion technique. The algorithm combines the popular pyramid-based and similarity-based fusion algorithms with the CNN model to provide a better fusion approach

Keywords :

Author : A.Yohan, Mr.P. Surya, M.Venkatesh, K.Nithin Reddy, K.Bala Adharsh,Y.Jagan Mohan Reddy

Title : DEEP LEARNING-DRIVEN MULTI-MODAL MEDICAL IMAGE FUSION USING CNN FOR ENHANCED BRAIN TUMOR DIAGNOSIS

Volume/Issue : 2026;03(02)

Page No : 46-51

A vital prerequisite for healthcare services is ensuring blood safety detection prior to transfusion. Transfusion hazards might arise from the presence of contaminants, clots, or aberrant aggregation in blood samples, even when blood group compatibility is proven. The manual microscopic observation used in conventional inspection techniques is labour-intensive. This paper presents an automated method for analysing transfusion compatibility using image processing techniques. The suggested method uses segmentation, noise filtering, grayscale conversion, and Canny edge recognition to handle microscopic blood pictures. To find anomalous structure patterns, morphological operations and numerical edge feature extraction are employed. The system decides if the blood sample is appropriate for transfusion based on the extracted attributes. For medical laboratories, this approach offers a quicker and more dependable initial screening tool while minimizing human participation.

Keywords : Blood Sample Analysis, Image Processing, Edge Detection, Medical Imaging, MATLAB, Automation

Author : Y. PRIYANKA, V.JAI KUMAR, V. KANCHANA, D. POOJITHA, M.V. MANASA

Title : BLOOD TRANSFUSION SUITABILITY DETECTION USING IMAGE PROCESSING

Volume/Issue : 2026;03(02)

Page No : 52-58

In this study, we created a unique modular neural network for precise human lung cancer diagnosis. we have taken the MRI reports of the patients and then we have evaluated the MRI images with the use of image processing techniques and neural networks to check whether the patient has been impacted by lung cancer or not. In order to improvise the MRI images for analysis we are grayscale function for making the images fit for the lung cancer. In order to determine the contrast and energy of the image, which are crucial factors in identifying lung cancer, we have applied the neural fuzzy classifications method. We have used the feature extraction approach to determine the image's entropy. We can determine whether or not the patient has lung cancer by getting the values of entropy, contrast, and energy.

Keywords : X-Ray CT Scan MRI Images, Artificial neural networks (ANN),Machine Learning, Lung cancer Detection, Medical Report Analysis

Author : P. GOVARDHAN, P. SIVA KRISHNA, P. PAVAN SRINIVAS, T. JAGADEESH, K. MANIDEEP

Title : DETECTION OF LUNG CANCER BY USING ARTIFICIAL NEURAL NETWORKS

Volume/Issue : 2026;03(02)

Page No : 59-64

Controlling electronic appliances in our homes is traditionally done through manual switches; however, with advancements in Internet of Things (IoT) technology, it is now possible to operate home appliances remotely from anywhere in the world. This project presents a smart home automation system integrated with a fish feeding mechanism that allows users to control and monitor household devices through a smartphone application. The system enables seamless operation of appliances with a single touch, providing convenience, flexibility, and enhanced control. The proposed solution is particularly beneficial for elderly and physically disabled individuals who may face difficulties in moving around and operating household devices manually. People living alone can also greatly benefit from such automation, as it reduces dependency on external assistance. In busy lifestyles, providing constant support to such individuals becomes challenging. Moreover, attempting to operate appliances manually may increase the risk of falls or injuries, which could lead to serious health consequences. By automating home appliances and essential tasks like fish feeding, the system minimizes physical effort, enhances safety, and improves overall quality of life through smart and efficient home management.

Keywords :

Author : Mr. A. GOPI, Mr. J. KANTHAIAH, K. GOPI CHAND,P. SRILAXMI, K. NARASIMHA RAO, P.SAI DEEPAK

Title : VOICE CONTROLLED HOME AUTOMATION USING MOBILE

Volume/Issue : 2026;03(02)

Page No : 65-72