TIET Digital Repository Thapar Institute of Engineering & Technology
Not a member yet
6889 research outputs found
Sort by
Gait Analysis and Estimation of GRF Towards Control of Lower-Limb Exoskeleton
No full textThis thesis presents the development of a comprehensive lower-limb exoskeleton framework aimed at augmenting human mobility and physical performance. A detailed kinematic analysis was carried out to define joint trajectories and limb configurations, employing Denavit-Hartenberg (D-H) approach suited for human-exoskeleton alignment. To validate the model and ensure its biomechanical relevance, experimental gait data was obtained using Gait Lab systems under IOR (Istituto Ortopedico Rizzoli) method for human motion capture protocols.
Further, Ground Reaction Force (GRF) estimation was performed as a critical element in kinematic modeling, allowing the system to reflect realistic load-bearing conditions during walking. Unlike purely analytical approaches, a learning-based estimation method was adopted for more accurate and adaptive results. The control techniques help in controlling the system and getting desired motion. To enhance system efficiency and effectiveness, advanced optimization techniques like GA, PSO and GA-PSO were implemented to fine-tune control parameters and ensure minimal tracking error with robust performance.
The integrated approach of combining biomechanics, gait analysis and control optimization culminates in a scalable and adaptive solution for human-exoskeleton synergy, laying the foundation for future development in assistive and rehabilitative robotic systems.DST/CRG project (CRG/2022/005578) for AMTI force plate and staff salar
Bio-Inspired Optimization Algorithms for Image Steganographic and Cryptographic Applications
No full textDepartment of Electronics and CommunicationThe multimedia data communication in the form of images is done in different applications, namely, healthcare, defense, and satellite. These images are prone to numerous attacks on the communication channel, such as statistical attacks and differential attacks. These attacks are overcome by adding the security layer to the multimedia data using steganography, encryption, and hybrid methods based on it. The steganography methods hide the secret data in the cover media, whereas encryption methods scramble the secret data using a random key. Further, in the hybrid methods, the secret data is encrypted and then hidden in the cover image. Hence, in this research, these security methods are investigated to find out the challenges of the existing algorithms and enhance them using the bio-inspired algorithms. In the literature, numerous authors used the bioinspired algorithm to enhance the security methods. However, in the literature, several bio-inspired algorithms are available, and selection of the most optimum algorithm is a challenging task. In this research, the most optimal bio-inspired algorithms, namely, Egyptian Vulture Optimization (EVO), Green Heron Optimization (GHO), and Black Widow Optimization (BWO), are chosen due to their better exploration rate, minimum parameter tuning, and better convergence rate. Based on these algorithms, security methods in the fields of steganography, encryption, and hybrid approaches based on them are designed in this research.
In this research work, two security methods are designed to enhance the security parameter known as imperceptibility for image steganography applications. In the first method, three parameters are determined before data hiding, such as optimal cover image index, block index, and secret data index, using the bio-inspired algorithms (EVO, GHO, and BWO), which were not claimed by other authors in the previous studies. On the other hand, in the second method, the secret data bits are matched with cover image LSB bits, and the matched index is determined. Thereafter, the matched index is hidden in the same cover image in the optimal way using the bio-inspired algorithm by finding the best starting index in the cover image and optimal secret data index. The bio-inspired algorithm searches the best indexes based on the objective function. In our work, the parameter mean square error (MSE) is taken as the objective function. Further, the benefit of the proposed image steganography method is that it is suitable for single- and multi-bit data embedding. The simulation evaluation of the image steganography method is done on the standard USC SIPI Image Database images. Further, for the evaluation purposes, several grayscale and color images are taken into consideration. Next, the evaluation is done based on subjective and objective analysis. In the subjective analysis, based on the visual quality, original cover images and their histograms are compared with the output images known as stego images. On the other hand, in the objective analysis, several performance parameters, namely, mean square error (MSE), root mean square error (RMSE), peak signal to noise ratio (PSNR), structure similarity index measure (SSIM), correlation coefficient (CC), entropy, university image quality (UIQ) index, image fidelity (IF), and normalized absolute error (NAE), are used to analyze the characteristics of the stego image with respect to the cover image. The result shows that the proposed image steganography method achieves high SSIM, CC, UIQI, IF near to one value, low MSE, RMSE, NAE near to zero value, and approximates similar entropy between cover and stego image as required in the steganography. Besides that, the proposed method achieves better PSNR without degrading the payload capacity as compared to existing methods.
Two security methods are proposed next to overcome the statistical and differential attacks on the secret data for image encryption applications. In the first method, a random key of 512- bits is generated using the bio-inspired BWO algorithm for data encryption. The benefit of the BWO algorithm is that it searches for the best key among the n number of keys based on the objective function. Subsequently, this key is utilized in the image encryption method to perform substitution, permutation, and key scheduling steps. Besides that, the BWO mutation operation is performed in the permutation step. On the other hand, in the second method, the BWO algorithm searches the best parameter values of the chaotic logistic map for key generation based on the objective function. After that, an exclusive-OR operation is performed between the secret image pixel and the random key. Next, the encrypted matrix is randomly circularly shifted horizontally and vertically to achieve permutation. In both methods, entropy is taken as the objective function. The simulation evaluation is done in the standard USC SIPI image database. In the evaluation, several images are taken into consideration. The result shows that the encrypted image is found to be completely noisy from visual analysis, and its histograms are equally distributed. On the other hand, in the objective analysis, the proposed methods achieve high entropy (~7.999) close to the ideal value and also high number of pixels change rates (NPCR) (~99%) with a low correlation coefficient (near to zero value) and PSNR (near to 8-12 dB). Further, comparative analysis shows that the proposed method outperforms in terms of entropy and NPCR over the existing methods.
Next, a privacy-preserving method is designed by hybridizing the image encryption and steganography methods. The novelty of the proposed privacy-preserving method is that the same evolutionary BWO algorithm is used for key generation, for secret data encryption, and for optimized data hiding. The benefit of the proposed method comes from the fact that the cover image plane chosen to hide the encrypted data is not fixed, as it is determined based on the pixel intensity value. Moreover, in the proposed method, only sensitive information about the user is encrypted. The visual analysis shows that the input and output images are similar, and the objective analysis shows that the cover plane and optimal starting pixel index are not static, achieving better PSNR (in the range of 54.1579-54.3132 dB), high CC, SSIM, IF, UIQI (near to 0.999 value), and similar entropy is obtained between input and output image. The proposed methods are useful for anyone who wants to communicate secret data in a more secure way.Visvesvaraya PhD Fellowship and MeitY through SMDP C2SD project
Design and Control of a Knee Exoskeleton with Multiple Actuators
No full textThis thesis presents the design, modelling and control of a knee exoskeleton aimed at assisting individuals with knee disorders during stand-to-sit-to-stand (STS) motions. The knee exoskeleton is developed by a four-bar mechanism actuated by a linear actuator and is controlled using electromyography (EMG) sensors to detect the user's muscle signals. The system is designed to provide the required knee joint torque and range of motion necessary for smooth transitions between standing and sitting postures. Through detailed simulations, it is demonstrated that the exoskeleton achieves the desired angle of thigh rotation and reduces user effort during STS motions. A comparative study between linear and rotary actuators reveals that both can provide adequate assistance at the knee joint, with the rotary actuator delivering a higher torque output. This feature allows the exoskeleton to be powered by either or by the actuator depending on the user's needs. To address reliability concerns, the exoskeleton incorporates a fault-tolerant control system using a fault detection, isolation and reconfiguration (FDI) technique. This system enables the exoskeleton to continue functioning even if one of the actuators experiences a fault, ensuring user safety and continuous operation. The design parameters of the knee exoskeleton are further optimized to enhance user comfort by maximizing the angle of thigh rotation during STS movements. The optimization is performed using the interior point method, implemented in MATLAB environment, to ensure optimal kinematic performance while maintaining mechanical feasibility. The bond graph technique is employed to model and simulate the system dynamics, offering an efficient framework for multi-domain system analysis and control.
Experimental validations confirm the effectiveness of the knee exoskeleton in reducing user effort during STS transitions. The exoskeleton provides up to 60% of the external assistance required, significantly easing the burden on the user. This work contributes to the advancement of knee exoskeletons by providing a robust, fault-tolerant design that can be used in rehabilitation settings or by individuals with mobility impairments. The results of this research pave the way for future developments in assistive devices designed to enhance mobility and independence for users with knee disorders.I would also like to acknowledge the FIST project to Department of Mechanical Engineering, TIET dated December 20, 2021. FIST Project No.: SR/FST/ET-II/2019/504(C)
Enhanced Lung Cancer Detection Using Advanced Deep Learning Techniques
No full textLung cancer is one of the leading causes of mortality worldwide, making early detection
and accurate classification crucial for effective treatment. This research investigates the use
of deep learning (DL) models for lung cancer detection through histopathological images.
The study looks at three main models Convolutional Neural Networks (CNNs), U-Net, and
Region-based CNN (R-CNN), along with different methods to extract features like Gabor
Filtering, Local Binary Pattern (LBP), Wavelet Transform, Scale-Invariant Feature Transform (SIFT), and Harris Corner Detection. A balanced dataset comprising 5,000 histopathological images-2,500 lung adenocarcinoma (lung-aca) and 2,500 normal lung (lung-n) images were used for classification.
Among the models tested, U-Net with LBP achieved the highest performance with 99.00%
accuracy, 98.00% sensitivity, 99.00% specificity, and an AUC of 0.99. U-Net also performed exceptionally with Wavelet Transform and SIFT, yielding accuracies of up to 98.00%
and an AUC of 1.00. CNNs showed good performance, particularly with Wavelet Transform
and SIFT 96.00% accuracy, while R-CNN demonstrated reasonable accuracy with Wavelet
Transform 93.00% and Gabor Filtering 91.00%. However, R-CNN struggled with Harris
Corner Detection, achieving only 53.00% accuracy.
This research emphasizes the importance of selecting appropriate feature extraction techniques to improve the accuracy of DL models in medical image analysis. While DL models
show great potential for lung cancer detection, challenges such as high computational demand, feature extraction inconsistencies, and model interpretability need to be addressed.
Future work may involve ensemble learning, 3D medical imaging, and advancements in
explainable AI to further enhance model performance and clinical applicability.
Keywords- Lung cancer Detection, Deep learning models, Histopathological image analysis
Study of alkali metal oxides doped magnesium vanadate glasses and glass ceramics
No full textThe present thesis is related to the study of (75) V2O5-(25-x) MgO-xA2O (A= Li, Na and K) (x =0-12) systems. These samples are synthesized by the quench technique followed by heat treatments to the selected samples. The physical, structural, thermal, and conducting properties of as prepared samples are studied using various experimental and testing techniques to determine their suitability as cathode material in battery and electrochemical fuel cells. The research work is presented in different chapters.
Chapter 1: This chapter describes the background of renewable and nonrenewable energy sources. Different energy sources are being explored that can meet the demand without harming the environment. Fuel cell and batteries could be a new sustainable and renewable source of energy. The battery has shown great potential to become an efficient and reliable energy source associated with its high efficiency, simple processing and environment friendly nature. The role of different components of the battery has also been discussed in this chapter. Since the present work is related to the development of cathode materials, therefore, the main focus has been given to the particularly cathode as described in context to specific properties such as thermal stability at battery operating temperature in oxidizing and reducing medium and their electrical conductivity.
Chapter 2: This chapter deals with the literature review on vanadium based glasses/glass ceramics/ceramics as cathode material for batteries, fuel cells and supercapacitors. The physical, structural, optical, thermal and conducting properties of the cathode materials are discussed in light of various processing parameters, dopants and their concentration. Based on the literature review the motivation of the present study along with the objectives of the present research is also given in the last of this chapter.
xviii
Chapter 3: This chapter gives details about the source of raw materials synthesis parameters and the experimental method employed for sample preparation and characterization to achieve the proposed research objectives. The technical details of the experiment are also given. It also explains the experimental conditions in which structural, optical, thermal and conducting analyses have been done. For phase determination and structural analysis XRD, FTIR, Raman spectroscopy was used. The thermal properties of as prepared samples were investigated using DTA and dilatometer. Microstructure and chemical analysis were done by FESEM and EDS analysis. The conductivity measurement was done on the impedance analyzer with AC biasing ±1V.
Chapter 4: This chapter is related to the results and discussion of the prepared sample as well as heat treated samples. In this chapter, interpretations of data obtained from various characterizations have been discussed. Two composition of (75)V2O5-(25)MgO (MV-25) and (60)V2O5-(40)MgO (MV-40) was prepared using the melt-quench technique. Differential thermal analysis (DTA) is applied to calculate the activation energy of crystallization along with other parameters using two different theoretical models. Higher activation energy is observed for partially crystallized MV-40 glass. A negative thermal expansion coefficient (NTEC) is observed in both glasses. However, MV-25 glass shows three distinct regions of NETC and these regions diminish as MgO content increases in the composition. These glasses may have applications in photonics where NTEC is required. The present finding would help design glasses and glass ceramics as cathodes for battery applications. Based on the findings MV-25 glasses were further investigated with alkali different dopants.
Thus, the composition of 75V2O5-(25-x)MgO-(x)Li2O (x= 0, 1.5, 3.0, 4.5, 6.0, 9.0, 12) is synthesized by the melt quench method. Effect of Li2O on devitrification physical, thermal,
xix
structural, and conducting properties of as-quenched samples are analyzed utilizing various experimental techniques. X-ray diffraction and DTA confirmed the formation of phase-separated glasses up to 3.0 mol% of Li2O. Above this concentration of Li2O, the samples are glass ceramic. With the increase in the concentration of Li2O, the density increases in all the samples. Raman spectra demonstrate that as the concentration of Li2O increases, there is a transition from VO5 units into different structural units such as VO4, VO3, and VO2 of vanadium oxide. The highest conductivity of x = 4.5 is observed i.e. 10-4 S/cm at 250ºC. The activation energy indicated that the present samples could be mixed conductors in nature. These samples may be used as cathode materials in energy storage devices due to their mixed conduction with an appropriate conductivity at 250ºC.
The composition of 75V2O5-(25-x) MgO-(x)Na2O with x=0, 1.5, 3.0, 4.5, 6.0, 9.0, and 12 are synthesized by melt quench technique. The prepared samples were characterized by the XRD method. The addition of Na2O converted glass into glass ceramic/ceramics. The density shows no trend with Na2O content in the glass compositions. DTA confirmed the glassy nature of the x=0 sample. For all samples, the conductivity variation as a function of the temperature follows an Arrhenius relationship. The highest conductivity is found for the x=12 sample i.e. 10-2 S/cm at 250°C with better thermal stability. The developed samples can find applications as cathode materials in Mg or Na-based batteries due to good conductivity with better thermal and structural stability.
Glasses are synthesized by melt quench technique using the composition of 75V2O5-(25-x) MgO-(x)K2O with (x= 0, 6.0, 9.0, 12, and 15). Synthesized samples were characterized by the XRD method. The concentration of K2O leads to phases separated glass formation as confirmed by differential scanning calorimeter (DSC) and FESEM. Broad bands in Raman also confirm the
xx
glass formation. For all the compositions the conductivity variation as the function of the temperature follows the Arrhenius equation. Further activation energy of conduction is also calculated using an Arrhenius relationship. The conductivity of phase-separated glasses increases with temperature. It was around 10-3 Scm-1 at 250°C which is in the required range for cathode material. Based on the results, it can be concluded that these materials are appropriate for batteries and electrochemical fuel cells, thermo electrical materials, switching devices, sensors, etc. Further three samples viz. MVL-12, MVN-12, and MVK-12 are heat treated and investigated for conductivity and electrochemical properties. Heat treated MVL-12 and MVN-12 samples exhibit excellent results and could be used as supercapacitors.
Chapter 5: The summary of the results is given in this chapter. The best conductivity is observed for the MVN-12 sample at 250°C~ 5.31×10-2 Scm-1. Sample MVN-12 shows excellent specific capacitance Cs ~ 469.26 (F/g) at a current density of 1 A/g. However, MVL-12 heat treated at 500 °C shows high capacitance retention of about 128.9% with > 140% columbic efficiency even after 2000 cycles. At the end of this chapter, a future scope of the current study is also provided
Synthesis and Application of Sawdust-derived Nanocellulose for Sustainable Energy Harvesting
No full textCellulose is the most abundant, biodegradable, and versatile material which are easily available
in nature and is used as a biomaterial for various applications. To increase the applicability of
cellulose, it’s mechanical, thermal, optical, and electrochemical properties are improved by
converting it into nanocellulose. Moreover, nanocellulose, when derived from waste residues,
not only provides materials for advanced applications (energy harvesting, environmental
remediation, sensing, and storage, etc.) but also a sustainable solution to treat waste build-up in
the environment. Herein, the waste sawdust is used to synthesize cellulose through alkaline
hydrolysis, followed by bleaching treatment, which is further, converted into nanocellulose (NC)
via physical, chemical, and biological treatments for energy harvesting. The structural and
morphological analysis of synthesized cellulose and NC is done by XRD, FTIR, and SEM. The
stability and particle size have been characterized by zeta potential and DLS. Next, the
synthesized NC has been combined with corn starch and PVA in different ratios to fabricate a
film by solvent casting, which is further characterized by using FT-IR, dielectric studies, and
triboelectric studies. Further, the nanocellulose-based triboelectric nanogenerators have been
designed by pairing the various composites as positive triboelectric material with the fluorinated
ethylene propylene (FEP) film as the negative triboelectric material. The optimized 1 wt % NCTENG exhibits an open circuit voltage of 1kV, short-circuit current of 37 µA, and a maximum
power density of 4.6 W/m2
at 107Ω of load resistance. This study facilitates the sustainable and
adaptable fabrication of TENGs for many electrochemical applications, such as energy
harvesting and sensors
The Relationship between Curiosity, Materialism and Impulse Buying: A Gender Based Study
No full textBackground and aim: With the increase in prevalence of consumerism, it has
become crucial to understand the psychological factors, such as curiosity and
materialism, which have an influence on the impulsive buying habits. The
purpose of this research is to examine the connection between these factors with
a focus on gender differences.
Methodology: Participants (N=150) were taken from both genders (M=75,
F=75). Tools: For studying the variable of curiosity, the Curiosity and
Exploration-Inventory II (CE-II), for the variable of materialism, the Material
Value Scale (MVS), and for the impulse Buying variable, the Buying
Impulsiveness Scale (BIS) were used.
Results and Conclusion: Results show a strong relationship between
materialism and impulse buying, while curiosity indicates a weaker association
with impulse buying. There were no discernible gender differences in impulse
buying behaviour. But gender differences were observed in the levels of curiosity
and materialism. The findings showed materialism to be a stronger predictor of
impulse buying behaviour than curiosity
Diversity of Arbuscular Mycorrhizal Fungi in seleniferous soils and their role in selenium sequestration and biotransformation
No full textPh.D. thesisThe present research aimed to investigate the molecular diversity of Arbuscular Mycorrhizal Fungi associated with plants growing in natural seleniferous soils and their role in protecting plants from Se toxicity. Arbuscular mycorrhizal fungal spores were isolated from natural seleniferous soil. Their morphological identification was done and the role of AMF inoculation to host plants in Se stress amelioration, plant growth promotion and improvement in bioactive and antioxidant potential was studied. Further, Se uptake and biotransformation in various plant parts of maize were studied via nursery trial and field experiment in natural selenium contaminated soil. For molecular identification, the genomic DNA extracted from maize roots grown in seleniferous and non-seleniferous regions was amplified using AMF specific primers by nested PCR approach. The 1.5 kb amplicon spanning pSSU-ITS-pLSU of 18S rRNA of AMF was deciphered using Illumina MiSeq Next Generation Sequencing (NGS) technique. A total of 17 AMF species from the seleniferous region and 18 AMF species from the non-seleniferous region were identified. The number of reads of Glomus irregularis, Glomus custos and Glomus intraradices was higher in seleniferous soil than non-seleniferous soil. A consortium of Se tolerant AMF inoculum was prepared and inoculated to maize seeds to be grown in seleniferous soils. AMF-inoculated plants were in good shape with higher root, shoot and grain biomass than non-AMF-inoculated plants. AMF inoculation lead to higher Se accumulation in roots but lesser Se accumulation in shoots and seeds of inoculated maize plants as compared to control plants.
Nursery trial in seleniferous soil revealed that AMF inoculated plants had 2.3-times, 2.3-times and 1.9-times higher root, shoot and maize cob biomass respectively as compared to control plants. Se uptake studies through fluorescence spectrometry revealed that AMF inoculation led to 1.1 times higher Se accumulation in mycorrhizal maize roots as compared to control roots, but lesser translocation to shoots and seeds i.e. 0.3-times and 0.5-times lesser.
Evaluation of growth parameters after field trial revealed that the root length and shoot length of AMF plants were of 2.13 times and 1.92 times respectively that of control plants. There was an upsurge of 2.35 folds, 2.32 folds and 1.93 folds in root biomass, shoot biomass and maize cob biomass respectively in AMF plants as compared to control plants. Bioaccumulation studies revealed that there was 7.80% higher Se accumulation in AMF roots as compared to control roots but at the same time, there was 71.83% and 49.92% lesser Se accumulation in shoots and grains of AMF plants. AMF inoculation increased Total phenolic (1.6-times) and Total flavonoid content (2.3-times) in host maize plants as compared to control plants. Antioxidant studies revealed that AMF inoculation also led to significant rise in catalase (1.4 & 2.2-times), superoxide dismutase (1.7 & 2.0-times), ascorbate peroxidase (1.2 &1.4-times) enzyme activities and DPPH radical scavenging activities (1.3 & 1.3-times) respectively in both shoots and roots of AMF inoculated plants.
The study of selenium speciation in various plant tissues of AMF and control plants from field trial carried out using X-Ray absorption near edge spectroscopy technique revealed that the toxic inorganic selenium species present in control plant roots were replaced by reduced elemental selenium in AMF inoculated plants. Volatile organic forms of selenium i.e. dimethyl selenide and dimethyldiselenide present in higher proportion in AMF inoculated plant tissues, escape from the plant and act as active response of the inoculated plants to evade Se toxicity.
Present research results suggest that AMF impart phytostablization potential to inoculated maize plants which means that such plants can stabilize or hold Se inside its roots along with its reduced translocation to shoots and grains. AMF further protect plants from selenium toxicity by biotransformation of Se to methylated volatile Se derivatives and improve plant growth promotion in maize plants by lesser Se translocation to above ground tissues while thriving under selenium stressed environment
Influence of Affective State Priming on Effortful Decision Making: Image versus Words
No full textUnderstanding how affective states have an impact on effortful decision making is crucial for the advancement of cognitive and applied psychology. Affect states are known to shape attention, memory,and other executive functions, yet their impact specifically on the perceived effort and willingness to engage on challenging or difficult tasks remain under-explored. Previous research has established that positive and negative affect state modulate cognitive flexibility and risk-taking, there is limited knowledge regarding how different modalities of affect (visual vs semantic)influence effort-based decisions, particularly in young adults.is thesis aims to investigate the effects of positive affect and its arousal levels- induced using visual (images) and semantic (words) stimuli- on effortful decision making. The study seeks to determine how these affective states shape judgments of task difficulty and willingness to exert effort. A within-subjects experimental decision was employed where the participants, of age range 18-25 (mean age ≈ 22), were exposed to standardized image and word stimuli using an Effort Expenditure for Rewards Task. Their subjective ratings for valence and intensity along with their choice and effort exerted in the task were recorded and analyzed. The results reveal that affective states, particularly negative affect, significantly reduced the participants willingness to choose high effort/ high reward tasks, while positive affect showed more flexible decision making. Additionally, the modality of the stimulus (images vs words) independently influenced the risk propensity and task performance, with images showing more affective responses than words. These results suggest that both affect and the modality play a critical roles in shaping how individuals approach effortful decision making