Kumaraguru College of Technology

Air pollution is a leading environmental health risk in India, contributing to disease burden and mortality. Effective mitigation measures require high-resolution monitoring. This study leverages hyperlocal air quality data from 21 low-cost sensor (LCS) stations in Bhubaneswar (March 2023–February 2024) to assess spatiotemporal variations in PM2.5. Additionally, the Multiple-Path Particle Dosimetry (MPPD) model was used to quantify wintertime PM2.5 deposition in the human respiratory tract across seven age groups. PM2.5 exhibited strong seasonal variation, peaking in winter and remaining elevated in post-monsoon. Persistent high-pollution regions, including Rasulgarh, Dumduma, Nandan Vihar, Patarapada experienced consistently high concentrations due to vehicular emissions, industrial activities, and biomass burning, emerging as persistent high-pollution regions necessitating year-round mitigation. Seasonal pollution hotspots - Kesora, Niladri Vihar, Sailashree Vihar experienced episodic PM2.5 spikes exceeding 20% of seasonal means, requiring targeted interventions. Dosimetry analysis revealed age-specific deposition patterns: infants and toddlers retained the highest PM2.5 in the head and tracheobronchial regions, increasing their risk of upper respiratory conditions, while children had the highest pulmonary deposition, posing long-term respiratory risks. Adults (> 49 years) exhibited lower pulmonary deposition but remained vulnerable to cumulative exposure effects. Lobar analysis showed predominant PM2.5 accumulation in the lower lung lobes across all age groups, with infants experiencing the highest deposition due to smaller airway diameters. These findings highlight the pressing need for targeted air pollution mitigation strategies in high-risk regions and among vulnerable populations to minimize long-term health impacts in the city.

A compact ultra-wide-stopband microwave bandpass filter is introduced in this work using an L-shaped quarter-wavelength resonator and stepped impedance resonators. The bandpass filter (BPF-I) is designed at 2.1 GHz (f) using an L-shaped quarter-wavelength line. The lowpass filter is developed at 4 GHz using stepped impedance resonators. The introduced ultra-wide-stopband filter is created through integrating the BPF-I and the lowpass filter. Filter size reduction is achieved through the L-shaped quarter-wavelength line, while the ultra-wide-stopband is attained through the stepped impedance resonators. The mathematical analysis of the introduced ultra-wide-stopband bandpass filter is conducted through the extraction of ABCD and S-parameters. The lumped elements equivalent circuits are realized for the introduced filter, and their S-parameters are plotted. The introduced filter is fabricated on an FR4 substrate, and its S-parameters are measured. The comparison of mathematical analysis, EM simulated, circuit simulated, and measured results, showing good agreement. Measurements show a 140.4% fractional bandwidth (FBW), better than 20 dB return loss, higher than 28 dB rejection up to 30 GHz (14.3f) in the out-of-band with many transmission zeros, and 0.8 dB insertion loss. 0.23g x 0.37g is the overall dimension of the introduced filter.

Microarray technology has transformed the biotechnological research to next level in the recent years. It provides the expression levels of various genes involved in a particular disease. Prostate cancer disease turned into life threatening cancer. The genes causing this disease are identified through the classification methods. These gene expression data have problems like high dimensional with low sample size which imposes active challenges in the existing classification algorithms. Feature selection techniques are applied in order to address the dimensionality issues. . This paper aims in analyzing the feature selection methods for classification of gene expression data of Prostate and identify the significant genes that have a major influence on the disease. The three different feature selection methods such as Filters, wrappers and embedded selectors are applied before the classification process for selecting the top ranked genes. Then, the extracted top ranked genes are applied on the classification algorithms such as SVM, k-NN, Random Forest and Artificial Neural Network. After the inclusion of feature selection technique, the classification accuracy is significantly boosted even with less number of genes. Random Forest classification algorithm outperforms other classification methods. The significant genes that has the major influence in prostate cancer disease are identified such as KLK3, GFI1, CXCR2 and TNFRSF10C.

This study addresses the challenges of energy management (EM) in alternating current (AC) microgrids (MGs) integrated with renewable energy sources (RESs), focusing on optimizing power balance, efficiency, and operational costs. Due to the alternation in renewable generation and demand, systems have to tackle the inefficiencies of power conversion and emissions related to the operation of backup generators. To overcome these issues, a novel hybrid approach, the Giant Trevally Optimizer-Self-Adaptive Physics-Informed Neural Network (GTO-SAPINN), is proposed. This approach aims to enhance system efficiency, minimize power loss, and reduce MG costs. In this method, the SAPINN forecasts demand and renewable generation patterns, ensuring stable energy supply. Meanwhile, GTO improves load balancing and distribution among RESs in AC MGs. The effectiveness of GTO-SAPINN is evaluated in MATLAB, compared against existing methods such as Beluga Whale Optimization, Flying Foxes Optimization-Deep Attention Dilated Residual Convolutional Neural Network, and Particle Swarm Optimization. Results reveal that GTO-SAPINN method achieves 99.1% efficiency at a total cost of €42 053, demonstrating superior cost-effectiveness and time efficiency over competing methods. This approach provides a promising, reliable solution for EM in AC MGs with RESs, optimizing energy distribution, and supporting sustainable MG operations.

Acoustics and thermal insulation materials are essential for improving energy efficiency and reducing environmental impact. This study explores the development of composite materials from waste cotton fiber, coffee husk, and sawdust, focusing on their thermal conductivity and sound insulation properties. Various sample compositions were prepared and tested for ceiling insulation applications, with weight ratios determined using a simplex lattice design. The results highlight the superior performance of these waste-derived composites compared to traditional insulation materials, offering a sustainable and effective alternative. The optimal composite composition, containing 33.33% cotton fiber, 33.33% coffee husk, and 33.33% sawdust, achieved the highest thermal insulation value of 0.052% and a thermal conductivity of 0.048 W/mK. Sound absorption coefficients (SAC) were measured using the impedance tube method (ASTM E1050) across frequencies from 1600 to 5000 Hz. The CFS4 composite demonstrated outstanding high-frequency sound absorption, particularly above 2500 Hz, while the increased thickness of the CFS6 composite enhanced sound absorption at medium and low frequencies. With sound absorption coefficients exceeding 82.0%, these materials exhibit exceptional acoustic properties. Moreover, thicker composites were found to improve thermal insulation significantly. These findings position the developed waste composites as a promising, eco-friendly solution for thermal and acoustic insulation in sustainable construction.

Small gas turbine (SGT) engines power the latest generation of unmanned aerial vehicles (UAVs). This study investigates the axial flow–natured rotor blades of an SGT compressor. Rotor structures in axial flow compressors have failed due to complex stress scenarios. Lightweight, high-resistance material can reduce the failure rate of axial flow compressors in fixed-wing UAVs. Compressor blade designs are completed in 3DEXPERIENCE after extensive material investigation, in which the design data are evaluated through standard analytical procedures. First, an axial flow compressor blade for a long-range UAV is analyzed using computational fluid dynamics, in which imposed the complicated rotodynamic conditions in a single moving reference frame to the axial flow compressor rotor blade. Curvature and proximity discretization are employed in conjunction with ANSYS Mesh in this study. The edges and faces are scaled by a local mesh facility. Because of the intricate nature of the flow, an enhanced wall treatment turbulence model based on the k-epsilon parameter was employed. The computational method based on fluid-structure interaction (FSI) uses aerodynamic pressure distributions on the compressor blade as a basis for structural analyses. For this structural investigation, FSI has been used with a single-coupling direction. Composite compressor blades are put through an FSI inspection to ensure they are perfect. Epoxy has been used to analyze the structural integrity of a variety of modern composites, including those made from carbon fiber–reinforced polymers, glass fiber–reinforced polymers, and Kevlar fiber–reinforced polymers. Researchers used ANSYS Workbench to conduct in-depth analyses of over 25 lightweight materials. The blades are free to move in reaction to a shift in position at low, medium, and high rotational speeds, while the main hub remains stationary. This study lays the door for the development of axial flow compressors using nontraditional lightweight high-resistance materials.

Electric drives play a crucial role in various industrial applications, requiring precise control and adaptability to changing parameters. This research proposes a hybrid approach combining Circle Search Algorithm (CSA) and Recalling-Enhanced Recurrent Neural Network (RERNN), termed CSA-RERNN, for intelligent controller-fed electric drives. The primary objective is to control the speed of electric drives using complex mechanical configurations and variable parameters. The CSA optimizes the voltage source inverter’s control signal, while RERNN predicts the optimal control signal. The system considered the variables of inertia, load torque, and rotor shaft angle. According to MATLAB/Simulink, the proposed method demonstrates superior performance compared to existing techniques, such as Singular Spectrum Analysis (SSA), Artificial Neural Network (ANN), and Bee Colony Optimization (BCO). The experimental results demonstrate that the CSA-RERNN approach achieves lower error rates and better speed control. The system maintains response characteristics despite changes in the moment of inertia of the object, thereby ensuring long-term performance stability without deactivating the adaptation mechanism. From the experiment it is evident that the system’s performance remains stable despite fluctuations in the moment of inertia, which changes between 1.2 kgm² and 2.3 kgm² over time. In addition, the system’s response characteristics, including speed and current waveforms, are consistent and stable. At 1976s, the reference angular speed (ωRef) reaches 0.9 rad/s, and the system adjusts accordingly. At the same time, the current varies, with the highest reading reaching 6 A and the lowest dipping to −6 A. Additionally, the proposed method proves more cost-effective than conventional intelligent controllers. Overall, the CSA-RERNN technique offers a robust and efficient solution for electric drive control in various industrial applications.

The energy management strategy (EMS) in a hybrid electric vehicle (HEV) is crucial for enhancing fuel economy and reducing pollutant emissions. Deep-CNN is a powerful tool for tasks like image recognition and classification. A Deep-convolutional neural network (CNN) is proposed to predict fuel consumption based on various driving conditions, such as acceleration, road gradient, and speed. To enhance the prediction performance of Deep-CNN, the weights are tuned optimally using a beta scented Dwarf Mongoose optimization (BS-DMO) Strategy leads to better performance including accurate fuel consumption prediction in Hybrid electric vehicles (HEVs). This optimization strategy focuses on improving performance by efficiently searching for the best parameters across different solutions. Accordingly, the proposed BS-DMO + CNN model demonstrated superior performance, achieving significantly lower error metrics, including mean absolute error (MAE), mean squared error (MSE), and root mean squared error (RMSE) compared to conventional methods. Sensitivity analysis confirms that an optimal parameter of 1.5 yields the best results, while convergence analysis shows that the BS-DMO + CNN consistently achieves minimal cost values, enhancing its effectiveness in energy management.

The widespread use of incense in indoor environments, particularly in cultural and religious practices, poses significant health risks due to particulate matter (PM) emissions. This study examines the chemical composition, particle morphology, and deposition dynamics of PM from four types of incense: Cup dhoop, Cone dhoop, Natural Incense Powder, and Agarbatti. Advanced analytical techniques, including SEM, FTIR, ICP-MS, and CAM, were employed to characterize particles, focusing on their size, elemental makeup, and surface properties. Particle sizes ranged from 12.02 µm to 422.3 nm, with lenses showing higher concentrations than filters. Elements such as sodium (300 µg/m³) and mercury (1.99 µg/m³) were prominent in lenses, while arsenic (6.2 µg/m³) and cadmium (0.19 µg/m³) were dominant in filters. Neurotoxins like aluminum, lead, and mercury highlighted potential risks, including oxidative stress and systemic toxicity. Deposition modeling revealed age-related differences, with children (8 years) experiencing higher pulmonary deposition (16.8% for Cup dhoop), while adults (21 years) showed greater head region deposition (37.6% for Agarbatti). Hydrophobic particles in filters (contact angle 119.2°) contrasted with hydrophilic particles in lenses (69.8°), increasing ocular exposure risks. Cone dhoop exhibited the highest cancer risk, affecting 5 in 100,000 individuals, emphasizing its hazardous nature. FTIR identified microplastics like polypropylene and polyvinyl chloride, known to adsorb and transport heavy metals, compounding health risks. These findings highlight the critical health impacts of incense emissions, particularly for children, and underscore the urgent need for stricter regulations, improved ventilation, and public awareness to mitigate exposure. Graphical Abstract

The electrodeposition method was used for FeNiWMoMn – High-entropy alloy (HEA) coating on a Copper substrate in an aqueous medium. The temperature was set at 75 °C and the deposition time was varied as 30 min, 60 min, and 90 min with a constant current density of 1A/dm². The 60-min deposited films were exposed to one hour of annealing at 200 °C to study the effects on their magnetic and structural properties. FeNiWMoMn alloy coatings were characterized using X-ray diffraction (XRD), Field Emission Scanning Electron Microscopy (FESEM), Energy Dispersive X-ray spectroscopy (EDS) and Electrochemical Impedance Spectroscopy (EIS). From the X-ray diffraction analysis, it was observed that they exhibit a cubic crystal structure. The crystalline size measured 27 nm, 26 nm, and 25 nm for deposition times of 30, 60, and 90 min, respectively. Annealed indicates increased crystallite size and reduced dislocation density, contributing to improved mechanical properties. The EDS results confirm that the sample has all of the required elements. The atomic weight percentage of Ni and Fe increases as the deposition period increases, whereas W and Mn decrease. The corrosion rate of coated FeNiWMoMn high-entropy alloy increases as the deposition time increases. The polarization resistance values start to decrease. After Annealing corrosion rate decreased and polarization resistance increased. The surface roughness properties of synthesized alloy are also investigated using AFM and found that the surface roughness of FeNiWMoMn alloy reduces as deposition time increases. Annealing improves the properties of Ni–Fe–W–Mo–Mn thin films for advanced applications.

This study presents a novel graphene-based terahertz (THz) absorber for carcinoma detection, utilizing a triangular split ring resonator (TSRR) loaded on a central cross dipole. Each arm of the cross dipole is embedded with a triangular SRR, enhancing electromagnetic confinement and optimizing absorption efficiency. The absorber is developed using polyimide as the dielectric substrate, while graphene forms the top frequency-sensitive surface, ensuring high absorption performance. The proposed structure exhibits near-perfect absorption at 0.8 THz, with an absorptivity close to 100%. The estimated theoretical bandwidth of the absorber is 20 GHz with reference absorptivity of 90%. Comprehensive simulations and parametric studies demonstrate polarization-independent behaviour under transverse electric (TE) and transverse magnetic (TM) modes. Additionally, the structure maintains angular stability up to 75°, attributed to the compact design and symmetric configuration of the TSRR-loaded unit cell. The absorber is further explored for biomedical sensing applications, where it is configured as a carcinoma sensor by introducing an analyte layer. The sensor exhibits a high resonance shift-based sensitivity of 287 GHz/RIU, making it a promising candidate for highly sensitive THz biosensing. The proposed design offers a compact and highly efficient platform for cancer diagnostics and biomedical applications in the THz regime.

Modern power distribution network incorporates distributed generation (DG) for numerous benefits. However, the incorporation creates numerous challenges in energy management and to handle the challenges it requires advanced optimization techniques for an effective operation of the network. Unlike traditional methods such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and standard Crow Search Algorithm (CSA), which suffer from premature convergence and limited adaptability to real-time variations, Reinforcement Learning Enhanced Crow Search Algorithm (RL-CSA) which is proposed in this research work solves network reconfiguration optimization problem and minimize energy losses. Unlike conventional heuristic methods, which follow predefined search patterns, RL-CSA dynamically refines its search trajectory based on real-time feedback, ensuring superior convergence speed and global search efficiency. The novel RL-CSA enables real-time adaptability and intelligent optimization for energy loss reduction in distributed networks. The proposed model validation is performed on the IEEE 33 and 69 Bus test systems considering diverse performance metrics such as power loss reduction, voltage stability, execution time, utilization efficiency for DG deployment, and energy cost minimization. Comparative results show that RL-CSA achieves a 78% reduction in energy losses, limiting power loss to 5 kW (IEEE 33-Bus) and 8 kW (IEEE 69-Bus) whereas traditional models converge at higher loss levels. The execution time is optimized to 1.4 s (IEEE 33-Bus) and 1.8 s (IEEE 69-Bus), significantly faster than GA, PSO, and CSA, making RL-CSA more efficient for real-time power distribution applications. By balancing exploration-exploitation using CSA while adapting search parameters through reinforcement learning, RL-CSA ensures scalability, improved DG utilization (98%), and better voltage stability (< 0.005 p.u.), making it a robust and intelligent alternative for modern smart grid optimization.

This paper aims to design a hybrid quadcopter that can be used for multiple detecting applications in which its performance parameters are studied under various maneuverings such as forward and vertical movements based on computational studies. In order to enhance the endurance, the conventional rectangular cross-sectional arm was replaced by airfoil cross sectional arm which helps in reduction of overall drag. The proposed idea is a combination of both tilt wing and tilt rotor configurations to the hybrid unmanned aerial vehicle (HUAV). The CAD modeling of UAV components such as wing and propeller is done using Autodesk Fusion 360 and the fluid flow analysis is carried out using ANSYS Workbench 23 software. Different test cases including the Computational fluid dynamics (CFD), Fluid structure interaction (FSI) analysis are executed to estimate the performance of the configuration. Analyzing from a stability point of view, a mathematical model was designed for control of altitude increment, hold and forward velocity accordingly, and tuning of the controller was taken over. This UAV is capable of attaining stability during harsh environments which is analyzed using control dynamics study and controller design processes executed. As a preliminary work for validation, grid convergence study is performed to obtain reliable outcome for the computational study taken over, in addition to the execution of analytical validation for estimation of aerodynamic forces and deflection of wing due to impingement of drag force over frontal area of wing and experimental validations to determine the thrust produced by propeller. The steady fluid flow analysis is carried over for wing planform and transient flow analysis is done for both vertical and forward propellers using advanced CFD techniques. Based on the FSI approach, structural analysis was carried over for wing and propeller through which the material selection was done. From which, GY-70-CFRP composite was concluded as the best performing material by analyzing the performance parameters including total deformation etc., among various different imposed materials based on aerodynamical loading. Interpreting from the performed analyses, the proposed configuration seems to operate at less power considering the lift forces induced, which also enables it to reach better altitudes at less RPM. The structural efficiency happens to be achieved due to the reduction in the RPM as there is a contribution in lift production as the angle of attack of the proposed wing increases, which also decreases rotors’ burden during forward motion and other maneuverings.

Hybrid renewable energy systems (HRES) provide clean energy, and promote a safe and hygienic environment by reducing greenhouse gas emissions. This paper proposes an HRES configuration that integrates with solar, wind, and tidal energy resources using Lyapunov optimization (LO-HES) with multi-objective parameters. LO-HES is implemented to manage real-time energy distribution with stability analysis. Simulation results indicate that in different environmental conditions and hybrid configurations, the suggested LO-HES achieve the best possible computation, maintaining system stability while optimizing the overall performance of all queues and improving energy utilization efficiency. In addition to that, extreme minimal load time and an optimal battery capacity of 680 kWh in diverse geographic regions, Energy efficiency of 89 % and low convergence time of 0.17 sec and cost reduction of 35 % to 40 %. The proposed system is particularly appropriate for quick application in rapidly changing environmental coastal conditions.

Nowadays, video monitoring applications have become significant for observing human activity with the help of computer vision-based approaches for investigating numerous video sequences. The major goal of anomaly identification is to discover the abnormalities automatically in a short interval of time. Performing efficient anomaly detection in a video monitoring system is considered as a complex task due to video noise, spilling, and anomalies. Various anomaly detection models based on Artificial Intelligence (AI) have been developed for video surveillance; however, these models often address only specific issues and do not consider the evaluation concerns over time. Hence, this paper aims to implement a video anomaly detection model through surveillance cameras for reducing abnormal activities that enhance the security of the environment. At first, the input videos are collected from the standard benchmark datasets. These collected videos are given in the frame extraction phase. Further, the extracted frames are fed to the object detection phase, where the YOLO-V3 technique is used. Parameter optimization of YOLO-V3 is achieved using the Modified Cat and Mouse Optimization (MCMO) algorithm to improve detection performance. The object-detected frames are fed as input to the ResNet for extracting the deep features. The extracted deep features are utilized for the classification phase, where the Optimized Bi-directional Long Short Term Memory (Bi-LSTM)-Radial Basis Function (RBF) (OBi-LSTM-RBF) provides the classified anomaly outcome. The variables are optimized using the enhanced CMO algorithm for enhancing the efficacy of the anomaly classification. Simulation evaluations are carried out to reveal the effectiveness of the offered approach with diverse baseline algorithms using diverse performance measures. The offered approach shows significant enhancement in accuracy over baseline approaches. Specifically, it outperforms conventional CNNs by 45.47%, DNNs by 90.03%, Bi-LSTMs by 34.6%, RBF by 90%, and Bi-LSTM -RBF by 89.7% at a learning percentage of 75. This enhancement in performance ensures the effectiveness of the recommended model in handling complex video data and detecting anomalies more precisely.

Poly Lactic Acid (PLA) is an environmentally friendly biopolymer widely used in bioimplants due to its biocompatibility, biodegradability, and favourable mechanical properties. Combining biocompatibility of PLA with ability of Digital Light Processing (DLP) technology to produce custom implants offers significant potential for advancing medical treatments and devices providing safer, more effective and tailored healthcare solutions. But, the research on impact of DLP parameters on mechanical properties of PLA is not well explored. The effect of DLP parameters namely Light Intensity, Orientation and Exposure time on tensile, flexural and impact properties were examined. UV and FTIR studies performed to analyse the binding energy and presence of functional groups in the utilized PLA. Intensity of light has greater impact on tensile (40%) and flexural strength (45.59%) whereas the impact strength is greatly influenced by the exposure time (41.05%). The adopted Machine Learning model predicts the mechanical properties with minimal error compared to regression models. The multi objective optimization performed through CRITICS-TODIM technique suggested 110% light intensity, 0° print orientation and 12 secs exposure time as the optimal condition.

Battery thermal management system (BTMS) is a very important field that is currently being focused on by the thermal and energy departments all around the world. This work primarily emphasizes channel design for BTMS and the utilization of modern computational fluid dynamics (CFD) investigations in BTMS. Enhancing the fluid‐battery heat transfer interaction is the aim of the proposed channel design. A reliable CFD study and better wall treatment confirmed the thermal performance of the identical channel design. Secondly, this study focuses on finding a suitable velocity at which a coolant can perform its best efficiently, that is, by absorbing most of the heat present in the battery system. Six coolant fluids were chosen to achieve the goal of finding the best velocity at three different heat generation rates (HGR). These HGRs include 5318, 19,452 and 42,400 W/m³ describing the C Ratings 1C, 2C and 3C, respectively. Six coolants were Ethylene Glycol, Propylene Glycol, Glycerine, Ethyl Alcohol, Water liquid and Water Glycol. It is concerning that even after choosing the required coolant for heat absorption, it becomes necessary that the velocity at which it can be allowed to flow through the battery system determines the effectiveness of the coolants. It was concluded that the coolant fluids better perform at 1 m/s. This lets us know that, when the flow of the coolant is at its lowest velocity, it can efficiently absorb the heat while it stays at that particular instant. The coolant's temperature was measured to be higher at the outlet (after it has flowed through the entire battery system) compared to the intake temperature. This indicates that the coolant has absorbed heat through molecular interaction. The input temperature was recorded at 29.85°C. It was also noted that Ethyl Alcohol and Propylene Glycol work the best at the HGR of 5318 W/m³, and the other coolants work the best at 19,452 W/m³ at 1 m/s. Using low‐velocity fluids in liquid BTMS has been found to enhance thermal management by improving heat transfer efficiency, ensuring structural integrity, extending the duration of heat exchange, enhancing temperature uniformity and reducing energy consumption. These factors collectively contribute to making lithium‐ion batteries safer and more effective for a range of applications.