M.K. Engineering College (RMKEC) is a private engineering college at Kavaraipettai, Gummidipoondi Taluk, Tamil Nadu, India, governed by the Lakshmikanthammal Educational Trust. It is affiliated to Anna University, Chennai, and accredited by All India Council for Technical Education(AICTE) with A+ Grade (3.52/4.00). All the seven departments of the college are accredited by the National Board of Accreditation(NBA). RMKEC is among the top engineering colleges of Anna University in Tamil Nadu and a Tier-I institution among self-financing colleges. It is a minority institution under language category.
In order to coordinate EVs along with renewable energy, it is necessary to have accurate forecasting, adaptive control, and a secure energy exchange. The hybrid framework that integrates Transformer forecasting with multi-agent reinforcement learning (MARL) suggested in this paper appears to highly suitable for the intended application. In fact, the Transformer encoder is the one responsible for the accurate predictions of EV load and renewable generation, while MARL policies give the required decentralization and dynamic coordination. Blockchain acts as a safety net for the transactions and a sign of trustworthiness for the prosumers, with IoT-level compression playing the role of latency eliminator in densely packed EV networks. Testing results show that predictive reliability is 96.9%, balancing is 32.7%, efficiency is 26.4%, and latency is 24 ms. Based on the comparison with ML baselines, the framework is 6.8% more accurate, 7.5% more balancing is achieved, 5.9% of the cost optimization is improved, and therefore, the EV–renewable integration is not only scalable but also resilient.
In this paper, an sustainable imperfect production system for shortages and complementary products with advertisement-price- sales return and green degree dependent demand and carbon emission is considered. The green production lot size, shortages and optimal pricing are three decision variables in third order equation. A provision has been made to sell the defective items. In literature, only two decision variables are considered in imperfect production system. In this model, three decision variables that is green production lot size, shortages and pricing are considered in imperfect production system. Price-break-even point is determined, the law of demand is verified, and highest possible profit is identified from the three alternative prices. The price break-even of the product per unit is 3299.4834 and in this stage neither profit nor loss for the concern. Taking from a third-order equation including three price variables, optimum pricing solution shows that a price level of 4079.5108 yields the highest profit of 102184.49. The objective of this model is to determine the green production lot size, shortages, and pricing for overall maximum profitability. The mathematical formulation's validity is shown by a numerical experiment, and the effects of inventory components on overall profit and managerial insights are examined using sensitivity analysis. The validation of the model's results was programmed utilizing Microsoft Visual Basic 6.0.
The untreated natural fiber reinforced polymer composites limited their broader structural applications due to their limited mechanical performance and poor interfacial adhesion. To counteract this, the tensile strength of modified banana fiber reinforced epoxy composites incorporated with nano-silica (SiO2) at different concentrations of 0.5 to 3 wt
Neurodegenerative diseases are characterized by gradual deterioration of the neurological system that impairs cognitive and physical skills. Early detection is critical to management. Neurodegenerative diseases provide obstacles due to their progressive nature, unknown origins and difficulty in early detection. In this Manuscript, early diagnosis of neurodegenerative diseases using temporal inductive path neural network in medical imaging (NDD-TIPNN) is proposed. Initially, the input images are collected from the Alzheimer's disease and Parkinson's Progression (ADPP) dataset. The collected images are then processed using Confidence Partitioning Sampling Filtering (CPSF) to remove image artifacts. The pre-processed image is subsequently fed into the Temporal Inductive Path Neural Network (TIPNN) for the classification of neurodegenerative diseases like Parkinson's, Alzheimer's and Healthy control. Generally, TIPNN does not show adopting optimization approaches to define the weight domains needed for accurate classification. Therefore, the Secretary Bird Optimization Algorithm (SBOA) is used to optimize the weight parameter of TIPNN which accurately classifies neurodegenerative diseases. The proposed NDD-TIPNN approach contains 99.19% higher accuracy, 99.43% higher sensitivity and 99.23% higher precision related to existing methods, like Spatially informed Bayesian neural network for neurodegenerative diseases classification (SBNN-NDC), Comparative analysis of machine learning algorithms for multi-syndrome classification of neurodegenerative syndromes (CNS-SVM) and Neurodegenerative diseases-Caps: a capsule network based early screening system for the classification of neurodegenerative diseases (CND-CNN) respectively. These results suggest that the proposed approach offers a highly effective and computationally efficient solution for the early diagnosis of neurodegenerative diseases, with a potential impact on clinical decision support systems.
Stroke is a type of cerebrovascular disorder that significantly affects a person’s health and well-being. Quantitative analysis of brain magnetic resonance imaging (MRI) images is crucial for evaluating and determining appropriate treatments for stroke. A blockage in the brain’s blood supply is a primary cause of stroke. To overcome these challenges, automated brain stroke detection in MRI using self-modulating convolutional neural networks for enhanced ischemic and hemorrhagic stroke classification (BSD-IHSC-SMCNN) is proposed. At first, the input image is gathered from the Brain Stroke MRI Images Dataset. Then, the input images are pre-processed utilizing a Bilinear Double-Order Filter (BDOF) is employed to enhance image quality and remove noise. Then, the pre-processed data are fed into a structured doubly stochastic graph based clustering (SDSGC), which is used to split the image into meaningful regions. Then, the segmented data are fed into feature extraction using the Spatial-Spectral Recurrent Transformer (SSRT). It is used to extract relevant features such as Stroke Area, Energy, Contrast, Homogeneity and Skewness. Finally, the Self-Modulating Convolutional Neural Network SMCNN is employed to detect and classify brain stroke into normal, ischemic, and hemorrhagic categories. To further develop classification performance, the Wombat Optimization Algorithm WOA is applied to optimize the network parameters and enhance overall accuracy. The BSD-IHSC-SMCNN model achieves the best performance in brain stroke detection, with 98% accuracy for Normal, 93% for Ischemic, and 96% for Hemorrhagic Stroke and 1.170 seconds computation time, with existing methods, such as the expression of SARS-CoV-2 spike protein in cerebral fluid, deep learning-driven multi-class classification of brain strokes using computed tomography: a step towards improved diagnostic precision (MCBS-CT-CNN), and a new deep learning (DL)-driven GUI design and implementation for automatic detection of brain strokes with CT images (BS-CT-ANN). The arteries: Hemorrhagic stroke implications HS-GAN post-mRNA vaccination.
