Kalyani Government Engineering College

The use of the oppositional crayfish optimization algorithm (OCOA) for the solution of electric vehicle charging station (EVCS) incorporation and network reconfiguration (NR) with distributed generation (DG) and capacitors placement problems in a radial distribution network (RDN), where active power loss and annual energy loss cost minimization are the main objectives of the study. An improved version of the crayfish optimization algorithm (COA) is created by adding oppositional behaviour into the primary COA algorithm for the generation of an opposite primary population to find the optimal solution. Two test networks (33-bus and 69-bus) are used to examine the effectiveness of the proposed OCOA algorithm with three different scenarios and they are (i) EVCS with unity power factor (UPF) DG and capacitor placement, (ii) EVCS with optimal power factor (OPF) DG and capacitor placement, and (iii) EVCS inclusion and network reconfiguration with optimal power factor (OPF) DG and capacitor placement. The OCOA method allows improvements of 81.45

Multimodal generative AI systems like Stable Diffusion, DALL-E, and MidJourney have fundamentally changed how synthetic images are created. These tools drive innovation but also enable the spread of misleading content, false information, and manipulated media. As generated images become harder to distinguish from photographs, detecting them has become an urgent priority. To combat this challenge, we release MS COCOAI, a novel dataset for AI generated image detection consisting of 96000 real and synthetic datapoints, built using the MS COCO dataset. To generate synthetic images, we use five generators: Stable Diffusion 3, Stable Diffusion 2.1, SDXL, DALL-E 3, and MidJourney v6. Based on the dataset, we propose two tasks: (1) classifying images as real or generated, and (2) identifying which model produced a given synthetic image. The dataset is available at https://huggingface.co/datasets/Rajarshi-Roy-research/Defactify_Image_Dataset.

This research presents a novel wide-range planar-type programmable optically variable resistor (POVR) utilizing TiO2 similar to epoxy nanocomposites. Conventional variable resistors, such as digital potentiometers and motorized rheostats, either lack continuous resistance modulation or exhibit cumbersome structures. Recent advancements using optically coupled light emitting diode (LED) - light dependent resistor (LDR) configurations improved automation capabilities but had limitations due to circular cross-sectional designs. Here, the authors propose planar-type POVRs with rectangular cross-sectional optical fiber spacers (ROFS), suitable for integration into modern planar electronic circuits. These ROFS cores were fabricated from epoxy resin incorporated with varying concentrations of titanium dioxide (TiO2) nanoparticles (0.0, 0.125, and 0.25 wt%) to precisely control optical and electrical properties. The resistance modulation was achieved by adjusting the forward bias voltages of red-green-blue (RGB)-LED sources, enabling continuous resistance variation from hundreds of Ohms to tens of mega-Ohms. Comprehensive analytical modelling was developed and validated through extensive experimental measurements, showing a close match with an error margin below 4 %. Additionally, detailed characterization, including field-emission scanning electron microscopy (FESEM), X-ray diffraction (XRD) analysis, and optical spectroscopy, confirmed uniform nanoparticle dispersion and favourable optical properties. The fabricated planar POVRs demonstrated excellent scalability, repeatability, and adaptability for advanced electronic applications.

The rapid growth of the Internet of Medical Things (IoMT) has transformed how healthcare data is collected and analyzed to improve patient health through real-time monitoring and personalized treatment. However, there is a significant concern over data security, privacy, and trust among healthcare providers. To address these challenges, we present a federated ledger and intelligence framework named FLIT to provide secure and privacy-preserving data sharing across multiple healthcare stakeholders. The proposed method combines blockchain technology with federated learning (FL) to protect sensitive patient information. Herein, patient data is never shared in raw form and is encrypted and distributed among trusted stakeholders so that individual data points remain confidential and tamper-proof. Besides, access control is enforced using smart contracts that automatically grant or deny permissions based on predefined conditions. To enhance clinical decision-making, FL allows hospitals and healthcare providers to collaboratively train machine learning models without exchanging actual patient data. These models are then used to generate recommendations for patients' treatment plans, which can be shared across the network. We also implemented and tested the FLIT using smart contracts deployed on Ethereum environments and FL models trained on real-world healthcare datasets. Our experiment results evaluated several metrics, such as local and global model accuracy, training loss, trust score, and cost of smart contract operations across multiple Ethereum Virtual Machines. The proposed work achieves high model performance and preserves privacy by validating its robustness in real-world healthcare scenarios.

This paper presents a new single-switch flyback converter with passive leakage recovery snubber comprising of three diodes, a clamp capacitor, and a small transformer, rated only 5