The global rise in energy demand, driven by modern lifestyles, necessitates more efficient wind energy harvesting. This research aims to determine the optimal blade angle for enhancing aerodynamic performance in Horizontal Axis Wind Turbines (HAWTs) under specific wind conditions. Computational and experimental analyses were conducted to evaluate lift and drag forces across different blade angles, focusing on maximizing moment and power output. The results identify 82° as the optimal blade angle for peak performance, with maximum moment observed at this angle for wind speeds of 3 m/s, 12.5 m/s, and 25 m/s. CFD simulations using ANSYS Fluent 17.0 with NACA aerofoils were validated through experiments, showing strong agreement between theoretical and experimental results. The study establishes a rated tip speed of 105 m/s for a 5 MW HAWT. By integrating experimental and computational approaches, this research provides valuable insights into the aerodynamic behavior of HAWTs, aiding in the development of efficient airfoils and blades to enhance energy generation. The findings highlight CFD's role as a cost-effective and time-efficient tool for assessing blade and wind characteristics, contributing to the optimization of wind energy systems and the advancement of sustainable energy solutions.
Smart transportation systems are easier to watch and make decisions about now that they have artificial intelligence in them. However, this has also made them much less secure, especially against hidden backdoor attacks. During training these attacks add poisoned samples with trigger patterns, which makes models produce outputs that the attacker controls when the trigger is present, but they still work normally on clean data. This research examines backdoor vulnerabilities in AI driven smart highway monitoring systems utilizing a traffic sign dataset. A trigger is added to a small part of the training data and then the data is given a new label for the target class. We trained an EfficientNet-B0 model on both clean and poisoned data, and then we test it using clean accuracy and attack success rate (ASR).The results shows a high clean accuracy of 99.84% and an ASR of 99.52%, which shows how stealthy and effective backdoor attacks can be. The results indicate that we need strong ways to find and protect against threats to make sure that AI is safe and reliable.
Soil improvement is considered an alternative when the natural soil is unable to meet engineering criteria. In the past, applying cement or lime to the soil was one of the most popular ways to improve it. Nowadays, a variety of modified soil improvement techniques are available. Certain treatments are not environmentally friendly, while others are challenging to implement. Bio-stabilisation techniques have emerged as a result of the development of alternatives to mechanical and chemical stabilisation for soil. The process of bio-stabilisation, which encourages ureolysis and results in the precipitation of calcite in the soil mass, commonly involves enzymes. According to recent research on environmentally friendly ground improvement methods, enzyme-induced calcite precipitation (EICP) is a suitable option for soil development. Soil sample was silty clay with liquid limit 69
VoiSOS is an application developed using Flutter which gets activated based on voice of the user. This application is initiated by a pre-trained voice command, which in turn initiates an SOS call automatically alerting predefined contacts. This work was all about developing a user-friendly interface, embedded with voice identification and automatic alert facility, tried for various test cases. Few challenges that we encountered were during the voice recognition optimization are in achieving optimal accuracy and false negative cases of initiating the alarm. When the situation is not feasible to make an emergency call, this work would be a definite saver effectively and efficiently. This module can also help the remote workers in enhancing their safety and security.
Consecutive k-out-of-n:G systems are widely used reliability models in engineering and communications, and their survival functions possess an inherent structural symmetry under the condition 2k≥n. This paper investigates the extropy characteristics of such systems under the condition 2k≥n, which admits a tractable closed-form representation of the survival function. A closed-form expression for the extropy of the system lifetime is derived using a probability integral transform and a density-quantile representation, which separates the symmetric structural contribution from the baseline distributional component. Bounds, stochastic ordering results based on a density-quantile order, and a characterization theorem are established. A nonparametric spacing-based estimator is proposed, its consistency is proved under explicit regularity conditions, and bootstrap confidence intervals are provided. Monte Carlo simulations under three component lifetime distributions (Weibull, Gamma, and Pareto II) demonstrate that the estimator is consistent, with bias and root mean squared error decreasing monotonically as sample size increases. A sensitivity analysis confirms the adequacy of the default window size rule. The proposed framework extends extropy to structured reliability systems and provides a practical nonparametric estimation tool with implementation guidelines, illustrated via a real data example on glass fiber breaking strengths.
