Sādhanā最新文献

This paper analyses the performance of multilevel half-bridge bidirectional DC–DC converter (MHBDC) under wide change in voltage transformation ratio. The advantage of MHBDC topology over conventional isolated bi-directional dc–dc converter (CIBDC) includes reduced voltage stress across the semiconductor switches and the transformer windings, which is suitable for high voltage applications in DC microgrids. Microgrids is inclined to voltage variations at both input and output side, which direct towards increased current stress in MHBDC. The rise in current stress can be minimized by employing various phase shift modulation techniques. The paper presents steady state time domain analysis of MHBDC under conventional phase shift modulation (CPSM), extended phase shift modulation (EPSM) and dual phase shift modulation (DPSM). A computational procedure is developed and demonstrated to obtain the phase shift angles at minimum value of current stress and varying voltage transformation ratio in the converter. Current stress analysis, soft switching region of operation and reverse power flow analysis in MHBDC are examined under different modulation schemes. A 1.5 kW hardware prototype of MHBDC is built, evaluated and presented the results. Power loss analysis and efficiency plots for the converter is determined and quantitatively evaluated with different modulation techniques considering high, medium and low power ranges.

Growing concerns about seismic events enforced structural engineers and architects to embrace the hazardous effect of ground motion in design. To address this, researchers have developed various base isolation (BI) techniques. This study comprehensively reviews BI system types, techniques, and implementation. Exploring the dynamic response of three-dimensional BI devices and the mutual effects of isolation devices and soil-structure interaction during strong ground motion, the paper covers topics such as seismic isolation of nuclear power plants, cost analysis, and various optimization techniques. Furthermore, the paper investigates the behavior of isolation devices in beyond-design events, including blast and aircraft impact loading. In general, the seismic isolation and control device response is demonstrated through shaking table tests and computational analysis. The study sheds light on the functions of seismic isolation system by comparing them with fixed base structures. Additionally, the paper presents codal recommendations, recent advancements, and current practices, aligning them with historical developments and past reviews of different BI techniques, along with their advantages and disadvantages. In conclusion, the closing remarks emphasize the future research prospects in this field.

The rotating vortex rope in the flow field generates low-frequency, high-amplitude pulsations at part load (PL) operation of a Francis turbine. These high amplitude pressure pulsations are responsible for power swing, cyclic fatigue, heavy structural vibrations, draft tube surge, and mechanical failure of turbine components. The fluid injection methods are generally employed to arrest or mitigate these low frequencies pressure pulsations. In the present work, the air is injected at a rate of 0.5% to 2.5% of the turbine flow rate at PL. The draft tube (DT) cone is modified to accommodate the air injection mechanism. Spectral analysis and spatial harmonic decomposition (SHD) are performed on pressure data for detailed investigation. The SHD is performed to get the synchronous (plunging component) and asynchronous fluctuations (rotating component) of pressure pulsations in the DT cone. The air injection in the DT cone successfully arrested the low-frequency pulsations of high amplitudes.

The collapse of shield tunnel seriously endangers the safety of urban subway construction and causes adverse social impact. In order to quantitatively evaluate the collapse risk of shield tunnel and ensure the normal construction of shield tunnel, the risk assessment model of shield tunnel collapse is established based on fuzzy theory and analytic hierarchy process. Taking the collapse risk assessment of the tunnel from Dahongmen South Station to Heyi Station of Beijing Metro Line 8 as an example, 4 first level indicators and 12 second level indicators are selected to define a fuzzy evaluation system. The weight set of evaluation factors is derived according to the hierarchy. According to the average membership degree of each index, the risk degree is calculated by fuzzy operator and the evaluation vector is obtained by applying the principle of multifactor fuzzy relation synthesis. The comparative analysis shows that the on-site monitoring results are very consistent with the lowrisk value obtained from the fuzzy evaluation, which proves that grouting reinforcement is the most effective method to prevent tunnel collapse. The engineering practice has verified the scientificity and rationality of this method, which gives full play to the advantages from qualitative to quantitative, and has high accuracy and universal applicability.

On a subcooled hydrophobic surface, the water vapor from humid air condenses into discrete liquid droplets and grows over time. The water vapor concentration field of the droplets in their immediate vicinity influences the growth of each droplet as the surrounding vapor field permeates the droplets. The diffusion of water vapor molecules in the humid air and their interaction with the liquid-vapor interface govern the growth of the individual condensate droplets. Nevertheless, most existing numerical models for condensation neglect the effect of neighboring droplets and the vapor concentration field surrounding each droplet. In this study, the point sink superposition method (PSSM) is used to calculate the growth rate of condensing droplets by approximating the inter-droplet interaction and the influences of the surrounding vapor concentration field. Further, this study simulated the condensation process on a vertically oriented hydrophobic surface by capturing condensation dynamics such as coalescence, droplet drainage, and renucleation.

Model-Free Control (MFC) is used to control a complex system by designing a simple representation of the system, known as ultralocal model. This MFC continuously updates the input-output behavior with the help of ultralocal model. The combination of MFC with standard Proportional Integral Derivative (PID) controller develops Intelligent Proportional Integral Derivative (i-PID) controller. The i-PID controller is basically a class of robust control technique in the field of PID controller. In this methodology, the tuning is quite straightforward for extremely nonlinear and/or time varying plants, without consideration of any modeling procedure. In Model Reference Adaptive Controller (MRAC)-PID controller principle, PID parameters are updated/tuned in accordance with control technique based on MRAC-Massachusetts Institute of Technology (MIT) rule, such that the plant is efficient to follow the reference model. The main objective of this paper is to implement MRAC based Intelligent PID controller for a class of dynamical systems using Model-Free Control technique, where the prior knowledge about the system dynamic is not essential and moreover complex parameter tuning is not necessary.

Research efforts have been centred on creating potential methods for producing green energy. A viable choice in this regard is fuel cells. Air breathing direct methanol fuel cell (AB-DMFC) is one of the many types of fuel cells, that is becoming increasing popular as a source of electricity for recharging portable electronic devices. The current study examined AB-DMFC performance equipped with various open ratio (OR)’s of cathode current collector (CC) such as 38.5%, 45.4%, and 55.4%. The cathode CC in an AB-DMFC is extremely vital to the operation of the fuel cell since it collects electrons, promotes oxygen flow for the cathode reaction and removes water bubbles produced at the cathode side, among other critical functions. For each open ratio of CC, the anodic fuel is supplied by external components like pumps with a concentration range of 1 M to 4 M methanol. When compared to other two open ratios, the experimental findings demonstrated that CC with a 45.4% OR at 2M of methanol concentration generated maximum power density (MPD) of 7.75 mW/cm². The next stage of experiments continued with the addition of alkali solution (NaOH) to methanol fuel, and it was found that adding alkali solution to methanol fuel enhanced the MPD of 8.5 mW/cm² and decreased methanol cross over compared with conventional AB-DMFC.

Intellectual property (IP) security has emerged as a critical concern in semiconductor industries. In the domain of hardware IP security, logic locking is a commonly used technique to prevent unauthorized access to IPs. This article proposes a conceptual pipeline to enhance the hardware IP security by leveraging generative models and logic locking concepts (GAN4IP) for hardware IP security. The proposed approach uses the concept of logic locking and generative adversarial networks (GANs) in a unified fashion to design secure hardware IPs. The GAN architecture uses deep learning techniques and graph-based representations of digital circuits to build obfuscated designs that can predict the behavior of locked netlists and generate secure designs. The proposed perspective method opens up new avenues for further investigation of highly secure electronic system design and has the potential to significantly impact the field of hardware IP security.

Myelin segmentation from real-time conventional MRI is a useful and challenging task in the medical field. In this work, myelin is segmented from real-time T1-weighted MRI after the application of suitable pre-processing methods. Myelin quantification from conventional magnetic resonance imaging (MRI) such as T1-weighted MRI is an innovative and challenging task in the research field. In the literature, no effort is accessible to segment and quantify the myelin from conventional MRI. Working with clinical data is an immensely demanding task. It is impossible to segment and quantify myelin directly from clinical data. It was necessary to employ pre-processing procedures before segmentation. In this task, Curvelet transform based adaptive Gaussian notch filtering with dynamic stretching is used before segmentation. Different image enhancement methods are compared, When compared to other image enhancement methods if we apply dynamic stretching then the segmented area is very nearer to the values calculated by the radiologists. The segmentation accuracy and other metrics also high in this case. The area from segmented myelin is calculated and the values are compared to the values calculated by the radiologist. The values are nearer to manual calculation in the case of Curvelet transform based adative Gaussian notch filtering with dynamic stretching.

Extracting keyphrases plays a vital role in the field of natural language processing, that focuses on recognizing and retrieving significant phrases that summarize the essential information in a document. This research paper introduces a novel approach to extract keyphrases using a statistical approach based on graphs that incorporates degree centrality, TextRank, closeness, and betweenness measures and natural language processing patterns. This approach involves constructing a graph representation of the document and identifying the most important nodes in the graph and leveraging natural language processing patterns to enhance the accuracy and relevance of the extracted keyphrases. The proposed model is examined on a standard dataset for performance evaluation and its outcomes are evaluated by comparing them with the state-of-art methods for extracting keyphrases. The precision, recall, and F-measure achieved by the proposed model are 0.5263, 0.5498, and 0.5323, respectively which shows that proposed model outperforms existing models. The principal novelty of this methodology resides in the utilization of statistical techniques based on graphs and patterns of natural language processing, which enable the detection of the most pertinent nodes and keyphrases of utmost significance. The proposed approach is generalizable to a wide range of domains and text types, making it a promising approach for keyphrase extraction in various applications, including content analysis, document classification, and search engine optimization. In conclusion, the proposed approach offers a robust and scalable solution for identifying keyphrases that capture the essential information of a document. Future research can build upon this approach to improve the efficiency and effectiveness of automated text analysis.