27
Arşivlenen Tez
0
DOI Atanmış
0%
DOI Oranı
Anabilim Dalı
Determination of antibacterial activity mechanism of red rose petal extract
Roses have been celebrated for centuries for their beauty, fragrance, and potential health benefits. Among the many bioactive compounds in red rose petals, anthocyanins stand out for their vibrant color and wide range of physiological effects. Anthocyanins, known for their antiatherogenic, anticancer, antidiabetic, anti-inflammatory, and antioxidant activities, are the primary pigments in red rose petals. Extracting these compounds can provide a visually appealing natural coloring agent suitable for various food products. Each component of a plant encompassing the trunk, bark, stems, leaves, fruits, roots, flowers, and seeds harbors a variety of chemical compounds, including phenolic acids, flavonoids, anthocyanins, and carotenoids. These compounds function as natural antioxidants, thereby mitigating the risk of plant diseases by delaying or preventing oxidative processes. This study focuses the antimicrobial properties of red rose petal extract and it is potential as natural preservative in the food industry. The extraction process involved freeze-drying the petals to preserve their bioactive constituents, followed by 70% ethanol extraction to obtain the desired extract. Frozen leaves were subjected to the vacuum freeze-dryer for a duration of 48 h. Moisture content was determined and the moisture content of the rose petal powder obtained in this study was found to be 4.65±0.05%. The antioxidant properties of red rose petal extract contribute to preserving food quality by preventing oxidative degradation and extending shelf life. Rose extract is used in medicinal chemistry to treat bacterial illnesses, targeting both Gram-positive and Gram-negative bacteria. This study assesses the antimicrobial activity and antioxidant potential of the extracted red rose petal extract. Through well diffusion assays and determination of Minimum Inhibitory Concentration (MIC) values, the study aims to identify susceptibilities of common foodborne pathogens against the extract as well as its antibacterial activity mechanism. The inherent antimicrobial properties of the rose extract position it as a promising candidate for utilization as a preservative in food products, thereby diminishing the risk of spoilage and ensuring food safety. Utilizing well diffusion assays, the efficacy of the extract against a spectrum of foodborne pathogens, including S. Typhimurium (31 mm), S. Enteritidis (23 mm), B. cereus (23 mm), S. aureus (25 mm), E. coli O157:H7 (23.5 mm), E. coli Biotype I (25mm), and L. monocytogenes (32.5 mm), was assessed. The discernible formation of zones of inhibition around the extract-treated discs signified inhibition of microbial growth, underscoring the extract's potential as an antimicrobial agent. Furthermore, MIC (Minimum Inhibition Concentration) values ranged from 114.07 mg/mL to 57.03 mg/mL. The determination of MIC values revealed S. Typhimurium and E. coli Biotype I had the highest MIC values and S. Enteritidis and B. cereus had the lowest MIC values. The petals of flowers are known as an ideal natural source of antioxidants from plant fiber due to their high flavonoid compound element level which includes anthocyanins.the red coloration of roses is attributed to the presence of carotenoids and anthocyanin. Plant based anthocyanins and carotenoids are valuable for food, nutrition, and pharmaceutical preparations due to low toxicity. The extract's total phenolic content (23.10 ± 0.81 mg GAE/g) and antioxidant activity were determined using established methods, including the FCR and DPPH assay, respectively. The observed antiradical activity ranged from 16.60 % to 74.15 %, indicating a dose-dependent response. The potential applications of red rose petal extract in food preservation are diverse. Its antioxidant activity can help retard lipid oxidation and preserve the quality of fats and oils in food products. Several studies were carried out to assess the time-dependent decrease in bacterial cell viability in the presence of rose petal extract. Initially, 24-hour bacterial cultures were prepared and supplemented with the rose petal extract, followed by incubation at 37°C to simulate optimal bacterial growth conditions. Sampling was conducted at predetermined intervals (0, 2, 4, 6, 24, and 48 hours) to monitor changes in bacterial cell viability over time. Each sampled culture was plated onto a Tryptic Soy Agar (TSA) medium to facilitate bacterial enumeration. A control group consisting of extract-free bacterial cultures was maintained to establish baseline bacterial growth patterns without the extract. The number of viable cells in the control tubes increased during the 24-hour incubation period for all bacterial cultures. In contrast, in the extract-added culture tubes, the number of viable cells reduced gradually, revealing the bactericidal effect of the red rose extract. A series of experiments assessed the potential damage to bacterial cytoplasmic membranes induced by rose petal extract. The extract was added to bacterial cultures at twice the Minimum Inhibitory Concentration (MIC × 2) and then incubated at 37°C for 24 hours. After the incubation period, the number of viable cells was determined by seeding onto two types of agar media: Tryptic Soy Agar (TSA) and TSA supplemented with 3% NaCl. Adding 3% NaCl to TSA created an osmotically stressful environment to challenge bacterial growth. Petri dishes containing the seeded agar were subsequently incubated at 37°C for 48 hours to facilitate colony formation. Following the incubation period, colonies were manually counted, and the difference in colony counts between TSA and TSA with 3% NaCl represented the number of cells with damaged membranes. The reduction percentages varied significantly among the tested strains: E. coli O157:H7 exhibited the highest reduction percentage at (80.83%), followed by S. Enteritidis at (64.52%), S. aureus at (34.51%), E. coli Biotype I at (14.86%), L. monocytogenes at (19.38%), S. Typhimurium at (11.5%) and B. cereus (6.72%). The investigation into the impact of rose petal extract on the tricarboxylic acid (TCA) cycle activity of pathogenic bacteria involved a series of experimental procedures. The extract was subsequently introduced into the suspension, adjusting bacterial cell concentration to 108 colony-forming units per ml (CFU/mL) during the logarithmic growth phase. Following a one-hour incubation period at 37°C, the mixture underwent centrifugation at 8000×g for ten minutes, forming a pellet. This pellet was then re- suspended in a 0.9% NaCl solution. Iodonitrotetrazolium chloride (INT) was then added to the suspension at a concentration of 1 mmol/L, and the mixture was further incubated for 30 minutes at 37°C. After the incubation, the maximum absorbance at 630 nm was measured using a spectrophotometer. By quantifying the maximum absorbance at 630 nm, changes in metabolic activity associated with the TCA cycle could be determined. Of note, the extract had a relatively minimal impact on the xxi metabolic activity of E. coli O157:H7. However, it demonstrated a more pronounced effect on the TCA cycle of Salmonella Typhimurium. Roses are emblematic of beauty, bravery, passion, and love, often revered as the king or queen of flowers. Their extensive range of applications offers significant advantages over most other flowers. The polyphenols from rose petals can enhance the value of by-products in fruit processing. Utilizing an enzyme-assisted extraction followed by spray drying, this green technology offers an eco-friendly alternative to traditional methods like organic solvent extraction. This natural oil and its constituents offer safe alternatives for incorporation into various applications within the food sector. Specifically, they may serve as additives in dietary supplements or functional food products. This study emphasizes the potential of red rose petal extract as a natural preservative and antimicrobial agent in food products. It highlights its promising applications in the food industry, suggesting it as a natural and health-promoting ingredient. Further research is needed to explore its effectiveness in different food matrices and ensure its safety for consumption, paving the way for its integration into various food products.
Modifiye riske maruz değer ve beklenen kayıp modellerinin geriye dönük testi: Lineer olmayan portföylerde uygulaması
The banks have to measure the market risk daily for the calculation of their capital adequacy. According to the Fundamental Review of Trading Book (FRTB) market risk revision, which was released in 2016 by the Basel Committee on Banking Supervision (BCBS), the expected shortfall (ES) will replace the value-at-risk (VaR) approach in order to capture the tail risks. In this paper, various risk management methodologies have been compared based on their performances using both the VaR and the ES. The data are based on three different currencies (USD/TRY, EUR/TRY, and EUR/USD) for the period from Jan 2nd, 2007 to Jan 4th, 2017. The methodologies have been applied to several portfolios of assets, ranging from a linear one (pure FX Position) to highly non-linear one (complex derivative securities on FX). The binomial backtest method is used for comparing backtesting performance and the empirical results indicate that the ES method, in lieu of the VaR methods, ensures the significant reduction in the capital adequacy for the semi-parametric models. In addition, the ES yields a considerable capital adequacy reduction compared to the VaR in linear portfolios. The reduction in loses strengths as the portfolios get more non-linear. These findings mainly highlight the importance of the convexity and the subadditivity features of the non-linear portfolios.
Hashgraph consensus algoritması kullanarak ıot cihazının sınıflandırılması için dağıtık karalisteleme protokolü
Industrial applications require highly reliable, secure, low-power and low-delay communications. However, wireless communication links in the industrial environment suffer from various channel impairments which can compromise these requirements. This thesis presents a new reliable blacklisting protocol for ensuring the Internet of Things (IoT) network security and mitigating the effects of interference caused by multipath Rayleigh fading using a distributed approach. The proposed blacklisting protocol is simulated over a distributed IoT network setup where flat Rayleigh fading disrupts Message Queuing Telemetry Transport (MQTT) communications. Distributed servers jointly decide in real-time whether to blacklist a device after individually performing anomaly detection and submitting their results to the hashgraph network. The IoT devices are classified by a device fingerprinting method using various machine learning (ML) algorithms that are trained with real-time packet capture data. The proposed blacklisting protocol is shown to increase the accuracy of blacklisting malignant devices from 42% to 82% as the number of servers increases from one to five for mixed attacks. It also achieves higher accuracies ranging between 47.2%-97.6% versus 47.4%-90.7% compared to the related work for Denial of Service (DoS) attacks. For this simulation with multiple servers, a novel solution for reducing the PER in a Rayleigh fading environment by searching optimal distributed server locations with a hybrid approach is also provided. The proposed protocol and the server location optimization algorithm are particularly suitable for the Industrial IoT (IIoT) in mitigating the effects of harsh communication environments in manufacturing facilities.
Veri akışı analizi kullanarak yeniden yapılandırılabilir CNN hızlandırıcı tasarımı
Dataflow reconfigurability plays a crucial role in Convolutional Neural Network (CNN) acceleration by determining the optimal dataflow pattern for convolution operations. Fully reconfigurable architectures provide versatility and high resource utilization by supporting multiple dataflow options, but this comes with increased design complexity and operational overhead. On the other hand, non-reconfigurable architectures, optimized for a single dataflow pattern, deliver high efficiency for specific tasks but lack adaptability. This thesis introduces a novel intermediate dataflow reconfigurable CNN accelerator that balances flexibility and efficiency by integrating key dataflow patterns, enhancing adaptability and performance across diverse CNN applications. Through a detailed analysis, key dataflows are identified, and a unique architectural unit is developed for dataflow selection, with an average of 0.15% excess latency compared to the optimal scenario. Our specialized systolic array architecture accommodates various kernel sizes, providing an additional layer of reconfigurability. Our architecture requires 39% less area and 35% less power than fully reconfigurable designs. Additionally, it delivers an average of 33% better performance compared to non-reconfigurable architectures. In terms of efficiency, it provides a 7% increase over fully reconfigurable designs and outperforms non-reconfigurable options by up to 3.57X.
Hareket sensörü veri dizilerinin görüntü temsilleri ve önceden eğitilmiş görme modelleri ile giyilebilir cihaz tabanlı kullanıcı kimlik tanıma
The common methods employed in User Identity Recognition (UIR) and verification are often vulnerable to cyber attacks, requiring more robust solutions. Motion sensor data and biometric data are used in tackling both the UIR and Human Activity Recognition (HAR) tasks. These tasks are mostly accomplished by using Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM), and CNN-LSTM hybrid models. We propose a method that employs pretrained CNN and vision transformer-based models to achieve the UIR task by classifying image representations of sensor data. We conduct a comparative study by evaluating the performance of various pretrained networks in the image classification task by processing four activity datasets comprising raw data sequences. We construct a new hybrid architecture which combines DeiT-B and DenseNet201 models in a parallel configuration. This study also compares two kinds of preprocessing methods which are spectrogram and wavelet spectrogram and introduces a novel approach that is fundamentally distinct from these methods. This technique fuses raw data, spectrogram, and wavelet spectrogram information. The DeiT-B model obtains the highest accuracy as 99.76% on the DSA Dataset; however, our new hybrid architecture that combines DeiT-B and DenseNet201 performs superior.
Enerji hasadı ve ortak enerji ve bilgi transferi için çalışma uzunluğu sınırlı kodları kullanarak kod dizaynı
Energy harvesting wireless networks and networks that benefit from wireless energy transfer have become popular in the last decade. In these networks, the users can obtain the required energy for transmission from an external source, which eliminates the need of battery replacement. Therefore, such networks have a high potential for applications in different areas including wireless sensor networks, wireless body networks and Internet of Things (IoT). While there have been many advancements for energy harvesting communications and joint energy and information transfer from information and communication theoretic perspectives in the literature, these subjects have not been studied from a practical coding and transmission point of view in depth. With the above motivation, in this thesis, we propose a serially concatenated coding scheme to communicate over binary energy harvesting communication channels with additive white Gaussian noise (AWGN), and design explicit and implementable codes for both long and short block lengths. Run length limited (RLL) codes are used to induce the required nonuniform input distributions for both cases. We employ low density parity check (LDPC) codes for long block lengths, while for short block length designs, we utilize convolutional codes for error correction. We consider different decoding approaches for the two cases, i.e., an iterative decoder is used for the former while Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm over the product trellis of the convolutional and run length limited codes is used for the latter. Also, by noticing that similar coding solutions can be employed, we extend our work to joint energy and information transfer for both scenarios. Numerical examples demonstrate that the newly optimized codes with an inner RLL code are superior to the point-to-point optimal codes for AWGN channels for long block lengths when energy harvesting or joint energy and information transfer is considered, and that, for the short block length case, concatenated convolutional and RLL codes with higher minimum distances offer excellent performance.
Çizge tabanlı okuma hizalandırması için dağıtık akıntı işleme sistemi
Optimized the sequence alignment pipelines are needed to minimize the time required to complete processing the short-read genomic data. Today there are many sequence alignment tools exist, yet few of them are capable of directly ingesting the streaming base-call data. The sequencing has to be entirely completed before the mainstream aligners can begin mapping the reads to the reference. The sequencing process can take days to complete. The output is then needs to be demultiplexed into individual reads and aligned to the reference, which can take several more hours. Overall time of a genomic analysis can be shortened significantly by progressively computing the alignments at the time when the reads are still being generated. It is important to have genomic analysis done as quickly as possible, especially in life critical situations. Here we introduce a distributed stream processing framework for aligning short-reads into a graph representation of the genome. The massively parallel nature of the genomic sequencing data requires a massively parallel computation architecture. Thus we have designed our pipeline called {\algname{}} to align many reads to a de Bruijn graph in parallel. Our aligning method is specialized for the sequencing technologies that are based on base-call cycles, such as produced by Illumina. The results are made available soon after the final bases from the sequencing devices has been emitted.
Ayrıştırılmış örtülü vektörlerle yüz özelliklerini düzenleme için resimden resime çeviri
We propose an image-to-image translation framework for facial attribute editing with disentangled interpretable latent directions. Facial attribute editing task faces the challenges of targeted attribute editing with controllable strength and disentanglement in the representations of attributes to preserve the other attributes during edits. For this goal, inspired by the latent space factorization works of fixed pretrained GANs, we design the attribute editing by latent space factorization, and for each attribute, we learn a linear direction that is orthogonal to the others. We train these directions with orthogonality constraints and disentanglement losses. To project images to semantically organized latent spaces, we set an encoder-decoder architecture with attention-based skip connections. We extensively compare with previous image translation algorithms and editing with pretrained GAN works. Our extensive experiments show that our method significantly improves over the state-of-the-arts.
Dikgen frekans bölmeli çoğullamaya dayalı akıllı radyo sistemlerinde varış zamanı kestirimi
Cognitive radio (CR) systems can efficiently utilize the radio spectrum due to their ability to sense environmental conditions and adapt their communications parameters (such as power,carrier frequency, and modulation) so as to enable dynamic reuse of the available spectrum. In this thesis, theoretical limits on time-of-arrival (TOA) estimation are derived for CRsystems in the presence of interference. Specifically, closed form expressions are obtained for Cramer-Rao bounds (CRBs) on TOA estimation in orthogonal frequency division multiplexing(OFDM) based CR systems in various scenarios. Based on the CRB expressions, an optimal power allocation strategy that provides the best possible TOA estimation accuracy is proposed. This strategy considers the constraints imposed by regulatory emission mask and the sensed interference spectrum.The maximum likelihood (ML) TOA estimator is derived for an OFDM-based signalling scheme, and its performance is investigated against the theoretical limits offered by the CRBexpressions. In addition, numerical results for the CRBs and ML TOA estimator are obtained and the effects of the optimal power allocation strategy on the accuracy of ML TOA estimator are examined in the absence / presence of interference. The use of optimal power allocation strategy insteadof the conventional power assignment scheme is demonstrated to provide significant gains in terms of the TOA estimation accuracy. Analysis of the performance sensitivity of the optimal powerallocation strategy to the uncertainty in spectrum estimation is performed, and the performance of optimal power allocation is observed to be consistently superior to that of the uniform powerallocation even for substantially high values of spectral estimation errors.
Derin öğrenme yöntemi ile MLO ve CC görüntüleri kullanılarak meme kanserinin sınıflandırılması
Bu çalışmada, meme kanseri teşhisi için popüler derin öğrenme modellerinden olan ResNet50 ve InceptionV3 mimarileri karşılaştırılmıştır. Kadınlar arasında en yaygın şekilde görülen malign neoplazilerden biri olan meme kanseri, zamanında tanı konulmadığında mortalite oranlarında ciddi artışlara neden olabilmektedir Bu nedenle, erken teşhis için kullanılan mamografik görüntülerin doğru bir şekilde değerlendirilmesi büyük önem taşımaktadır. Ancak mamografi görüntülerinin manuel olarak yorumlanması, çeşitli hata risklerini barındırmakta ve sonuçların doğruluğunu etkileyebilmektedir. Bu bağlamda, yapay zeka ve derin öğrenme tabanlı otomatik tanı sistemleri, tıbbi görüntülemede meme kanseri tespiti için önemli bir alternatif olarak öne çıkmaktadır. DDSM (Digital Database for Screening Mammography) veri seti kullanılarak yapılan çalışmada, her iki modelin mamografik görüntüler üzerinden meme kanseri sınıflandırma performansları incelenmiştir. Bu süreçte, hem Cranio-Caudal (CC) hem de Medio-Lateral Oblique (MLO) açılardan alınan mamografi görüntüleri kullanılarak, modellerin kategorik doğruluk metrikleri üzerinden karşılaştırmaları yapılmıştır. Elde edilen sonuçlar, ResNet50 ve InceptionV3 modellerinin meme kanseri sınıflandırmasında farklı performanslar sergilediğini göstermektedir. ResNet50 modeli özellikle CC görüntülerinde daha yüksek doğruluk oranları sunarken, InceptionV3 modeli hem CC hem de MLO görüntülerinde tutarlı ve rekabetçi bir performans sergilemiştir. Sonuçlar, her iki modelin de otomatik meme kanseri teşhisinde etkin bir şekilde kullanılabileceğini gösterirken, farklı veri türlerinde ve görüntü açılarında performans farklılıklarının olduğu görülmüştür. Bu çalışma, yapay zeka tabanlı otomatik tanı sistemlerinin meme kanseri teşhisinde kullanılabilirliğine dair önemli bulgular sunmakta ve bu sistemlerin klinik uygulamalarda yaygınlaştırılmasına yönelik potansiyel katkılar sağlamaktadır.
Güç sistemlerindeki bozulmanın tespiti için genlik ve frekans değerlendirmesine dayalı olarak yeni yöntemlerin geliştirilmesi
Enerji sektöründe rekabetin artması ve kullanıcıların daha kaliteli enerji talep etmesi, elektrik şebekelerinde güç kalitesini öncelikli bir konu haline getirmiştir. Elektrik şebekelerinde kullanılan enerjinin kaliteli olabilmesi için sürekli, belirli bir gerilim seviyesinde sabit frekanslı ve genlikli olması gerekir. Ancak bunlar normalde tam olarak sağlanmaz. Elektrik güç sistemlerinde gerilim-akım karakteristiği doğrusal olmayan elemanlar nedeniyle gerilim ve akım dalga şekli lineerlikten uzaklaşmaktadır. Günümüz modern güç sistemlerinde doğrusal olmayan yüklerin, güç elektroniği devre elemanlarının ve devrelerinin artmasıyla güç kalitesi problemleri de artmıştır. Güç kalitesini etkileyen unsurların başında harmonikler gelmektedir. Enerji nakil hatlarında lineer olmayan yükler, doğrultucular, ark fırınları, eviriciler gibi elemanlar harmonik üretirler. Harmoniklerin elektrik güç sistemleri üzerindeki etkileri ise; aşırı ısınmalar, gerilim düşümleri, ek kayıplar, rezonans olayları, dielektrik zorlanmaları, koruma ve kontrol yapan ölçme sistemlerinin hatalı çalışması gibi birçok istenmeyen durum oluşturmaktadır. Elektrik sisteminde, doğrusal olmayan yüklerin özellikle Elektrik Ark Ocaklarının çalışması nedeniyle, ortak bağlantı noktalarında akım ve gerilim kaynağının tespit edilmesi ve elektrik hatlarına ne kadar bileşen katkısı yapıldığının belirlenmesi büyük önem arz etmektedir. Bu nedenle harmoniklerin tespit edilmesi, ölçülmesi, kestirimi ve harmoniklerin bastırılması son derece önemlidir. Harmoniklerin genlik, frekans ve faz açısından doğru tahmini ve engellenmesi, güç kalitesini ve verimi artırır. Yapay zekâ ve akıllı sistemlerin gelişmesiyle harmoniklerin tahmini için Fourier dönüşümü temelli algoritmaların yanı sıra akıllı yöntemler de kullanılmaktadır. Bu tez çalışmasında harmoniklerin genlik ve faz kestirimi için küçük kareler yöntemi ile birçok meta sezgisel algoritmalardan oluşan hibrit yöntemler denenmiş ve geliştirilmiştir. Harmoniklerin genliği en küçük kareler yöntemiyle hesaplanırken faz açıları birçok optimizasyon algoritması ile tahmin edilmiştir. Önerilen algoritmalar literatürde önerilen test sinyallerinde ve gerçek veri setinde kullanılarak başarısı test edilmiştir. Güç sistemlerindeki harmonikler son yıllarda üzerinde sıkça çalışılan bir araştırma alanı olmuştur. Bu tez çalışmasında, literatürdeki harmonikli sinyaller üzerinde PSO (Particle Swarm Optimization) ve Genetik Algoritma ile ilk çalışmalara başlanmış olup, daha sonra birçok optimizasyon yöntemi denenmiştir. EOA (Election Based Optimization), GWO (Grey Wolf Optimizer) ve GOA (Grasshopper Optimization) kullanılarak farklı frekanslarda yöntemler denenmiştir. Sonrasında son yıllarda geliştirilen AVOA (African Vulture Optimization Algorithm), ARO (Artificial Rabbit Optimization), SWO (Spider Wasp Optimization), MGO (Mountain Gazelle Optimization) ve AO (Aquila Optimization) algoritmaları ile harmonik kestirimleri yapılmıştır. Analizlerde, literatürde sıkça kullanılan birkaç test sinyali üzerinde çalışılmıştır. Bu sinyalin harmonik genlikleri, En Küçük Kareler (Least Squares) yöntemiyle belirlenmiş, faz açıları ise ilgili meta sezgisel algoritmalar kullanılarak tahmin edilmiştir. IEEE 1159.2 Working Group kısmındaki File wave14a.xls gerçek veri seti üzerinde en iyi sonuç veren AO algoritması ile harmonik kestirimi başarı ile sağlanmıştır. Geliştirilen algoritmalar MATLAB yazılımı ortamında denenmiş ve sonuçlar benzer yapılmış çalışmalarla ve geçerlilik analizini yaptığımız tüm algoritmalarla karşılaştırılarak tartışılmıştır. Elde edilen sonuçlar, incelenen tüm yöntem tahminleri gürültüsüz ve gürültülü koşullarda bile doğru ve güvenilir harmonik tespiti yapıldığını göstermiştir. Harmoniklerin tespitinde, Aquila Optimizasyon algoritması ile En Küçük yönteminin entegre kullanımı, en doğru ve tutarlı sonuçları sağlamıştır. Önerilen AO-LS yöntemi, farklı test veri setleri üzerinde uygulanarak geçerliliği kanıtlanmıştır. Ayrıca, AO-LS yönteminin Benzetimli Tavlama (Simulated Annealing- SA) algoritması ile hibrit bir şekilde birleştirilmesiyle, harmonik tespit performansının daha da iyileştirildiği gözlemlenmiştir. Bu tez çalışmasında önerilen hibrit yöntem (SALSAO), sadece harmonik bozunumlarının tespitinde değil, aynı zamanda güç sistemlerindeki diğer bozunum türlerinin analizinde de test edilmiş başarılı sonuçlar elde edilmiştir. Yöntemin performansı hem sentetik test verileri hem de gerçek veriler üzerinde değerlendirilmiş ve gürültülü ve gürültüsüz koşullarda dahi yüksek doğruluk ve güvenilirlik sağladığı tespit edilmiştir. Elde edilen bulgular, SALSAO algoritmasının güç kalitesi problemlerinin çözümünde etkili bir yaklaşım olduğunu ortaya koymaktadır. ________________________________________
Kiriş aralığının öngerilmeli beton kiriş üzerine tabliyeli karayolu köprülerinin yapım maliyetine ve deprem performansına etkisi
This study examines the effect of using different girder spacing on the total bridge construction cost in varied seismic zones. For this purpose, a number of structural models are built utilizing the finite element analysis to study the superstructure and substructure of a benchmark bridge in detail. Using these models, related parametric analyses are conducted for altering girder spacing, span lengths, number of spans, column heights, soil types and seismic zones. Ninety-five bridges with distinct types of superstructures and substructures are then designed and analyzed. Finally, pertinent construction costs are estimated for each bridge model under consideration. Comparison of costs revealed that an increase in girder spacing leads to a decrease in the total bridge construction cost. Moreover, seismic performance analyses of the bridges showed that no considerable change in terms of seismic performance observed with the increase in the girder spacing.
Modifikasyon yapılmış düz dişliler için kontak analizi
Gears are subjected to high speed and high power especially in critical applications. It is known that these high power gears are mostly subjected to main failure modes due to contact stress, bending stress, and fatigue. In this thesis, a new analytical model is developed for contact stress in selected gear pairs. The new analytical method is built by combining the Weber Banaschek Method with the Sainsot Method for mesh stiffness and gear foundation stiffness, respectively. Results obtained from the newly developed method are compared to the results obtained from available transmission tools (MASTA and KISSsoft). Combination of these two analytical methods yielded accurate results. It is also seen that the analytical solutions are more close to KISSsoft results than the MASTA solutions. The contact stress calculations according to the new method resulted close to ISO standard and compared to the FEM module of MASTA. The results obtained from the new analytical method are close to experimental results by 96%. The MASTA Programme tends to give higher results on contact pressure compared to KISSsoft solutions. Finally, calculations according to AGMA 2001-C95 standard yielded in lower safety factors compared to the ISO 6336 standard.
Kemik doku mühendisliği için doğal kökenli silika takviyeli çift fiberli matrisler
Graft therapy is used to treat bone tissue loss, which has drawbacks; donor scarcity, risk of disease transmission and immune reaction. Tissue engineering scaffolds can overcome these drawbacks. In this study, a 3D scaffold that will support tissue regeneration at defect site was developed using mainly natural materials. Scaffold was produced by co-electrospinning and had two fiber phases; first phase was produced from a bacterial origin polymer, Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV), and small amount of Polycaprolactone to support fiber structure, second phase was formed by pullulan (PUL) and diatom silica shells (DS) from single cell algae origin. Also, in the study, PHBV production by Cupriavidus necator bacterial strain was optimized and used to produce scaffolds. PHBV/PCL/CA:PUL/DS scaffold group was produced as dual fiber matrix and new crosslinking method for PUL was developed for crosslinking through electrospinning. Cefuroxime Axetil antibiotic was loaded into hydrophobic PHBV fibers. Characterization studies revealed that, in aqueous environment scaffold degrade slowly, take less water and has stable structure, support controlled release of antibiotic and improved compressive strength due to incorporated DS and double fiber structure. In vitro studies revealed that DS bearing groups improved Saos-2 cell viability and co-electrospun groups that have hydrophilic PUL fibers supported L929 cell viability. Scanning electrone microscopy and confocal laser scanning microscopy analyses showed that cells distributed through co-electrospun scaffold with healthy morphology. In the study, a novel, 3D, dual fiber scaffold was produced with natural origin base materials and has potential to be used for bone tissue engineering applications.
Kat sayısı ve yükseklik-genişlik oranının deprem yalıtımlı binaların performansına etkisi
Despite the advanced engineering solutions, earthquakes remain to be one of the causes of casualties and property damage for humankind among natural disasters. To ensure that the essential services are not interrupted after an earthquake, it is crucial for some buildings to remain functional with minimal structural or nonstructural damage. These buildings include administrative buildings, law enforcement buildings, fire stations, hospitals, and communication centers. Seismic base isolation systems are frequently used for reducing quake-induced structural and nonstructural damage. This thesis mainly focuses on the feasibility of seismic base-isolated structures and determining the maximum feasible number of storeys of the base-isolated structures. Moreover, the effect of the following parameters on the seismic behavior is investigated: (i) isolator parameters, (ii) soil stiffness, (iii) earthquake intensities, and (iv) the aspect ratio of the structure. For this purpose, a comparative assessment is conducted to determine the feasibility of base-isolated structures. The feasibility of the base-isolated structures is assessed by considering the structural and nonstructural performance goals.
Uzak fay sismik bölgelerinde köprülerin performansa dayalı sismik tasarımı için deprem yükü azaltma katsayılarının geliştirilmesi
In this thesis, a methodology to develop strength reduction factors for performance-based seismic design of bridges in far-fault seismic regions is presented. The presented methodology is mainly based on performing linear 5%-damped response spectrum analyses (RSA) and nonlinear time history analyses (NTHA) of bridge piers modeled as single degree of freedom (SDOF) systems. Bridge piers of both circular and rectangular sections are analyzed for wide ranges of various design parameters considering several substructure-superstructure connections. Subsequently, a set of strength reduction factor (R-factor) equations derived by performing parametric linear regression analyses on the data resulted from the conducted structural analyses is proposed. The proposed R-factor equations are formulated in terms of the design parameters that significantly affect R-factor, and the maximum displacement ductility obtained from the NTHA considering various failure modes such as shear, shear-flexural, and flexural failure modes. Moreover, other equations to estimate the yield curvature, ultimate curvature, and yield moment capacities of bridge piers are proposed. Such equations that could be used to estimate the design displacement and flexural strength capacity of piers in the performance-based seismic design of bridges. Based on the derived equations in this thesis, a new performance-based seismic design procedure for bridge structures that mainly depends on R-factor is proposed. Finally, the implementation of the proposed design procedure is explained via design examples of bridge structures.
Deprem yalıtımlı binalarda eğri yüzeyli kayıcı yalıtım birimlerindeki tasarım sürtünme katsayısının gerçek dağılımı dikkate alınarak yapının burulma davranışının incelenmesi
In this study, torsional response of seismic isolated buildings with curved surface slider (CSS) isolators are investigated considering actual distribution of design coefficient of friction in the CSS isolator. For this purpose, experimental results are used to characterize frictional properties of CSS isolator. Then, nonlinear response history analyses are performed using three-dimensional model of 3 structures that are developed based on a hospital structure. Torsional response of the structures and the parameters that influence the torsion are investigated. The analyses revealed that if the effect of contact pressure and heating on the friction coefficient is not taken into account, the additional displacements in the CSS isolators due to torsion are significantly underestimated. It is also found that increasing the number of groups within isolation system is an effective way to reduce torsion, if the isolators are seperated into groups based on axial loads on them.
Deprem yalıtımlı binalar için eşdeğer doğrusal analiz yönteminin doğruluğunun değerlendirilmesi
Earthquakes have been the leading cause of loss of lives and properties since the beginning of human history. After major earthquakes, we see that even the buildings constructed using modern engineering solutions suffer severe damage or even collapse. Seismic base isolation systems are used to minimize earthquake-induced damage. This thesis focuses on evaluating the accuracy of the equivalent linear analysis method in the analysis of seismic base-isolated buildings. In order to do that, certain buildings with different story numbers, widths, and isolator parameters were analyzed by using a set of ground motions selected for different soil properties, and results were compared with the results obtained from the equivalent linear analysis method.
Çevik uçakların optimal girdi tasarımı ve sistem tanımlaması
This doctorate study aims to provide a methodology for developing aerodynamic and engine thrust models using simulated flight test data for the F16 fighter aircraft. An accurate and comprehensive representation of an aircraft's aerodynamic characteristics is required to design a flight control system or develop a high-fidelity flight simulator. Modern computational methods and wind tunnel testing can provide the aerodynamic database, but flight test data is required to obtain a more accurate and realistic aerodynamic database. As a result, system identification methods can characterize applied forces and moments acting on the aircraft. The F-16 nonlinear model also includes sensor models to simulate the actual flight data. The flight tests are carried out in the F16 simulation model using different excitations on the control surfaces. Simulation data is collected in predefined trim points. The equation error and output error methods are employed to analyze simulated data to estimate aerodynamic parameters in the time domain. The equation-error method is used firstly to identify aerodynamic parameters, and the results are then utilized as a starting point in the output-error process for fine-tuning. In general, thrust forces and moments are obtained from ground tests. The contribution of this doctoral study is to implement an iterative aerodynamic and thrust estimation approach in the absence of engine manufacturer data. The validation of resulting models is accomplished by comparing the measured flight data to the model's predictions for identical control inputs, as specified by the Federal Aviation Administration (FAA).
Diş doku mühendisliği uygulamalarına yönelik biyoaktif ajan yüklü hidrojel sistemleri
Use of tissue engineered oral restoration products is currently a popular approach for treating dental defects that adversely affect oral health in ageing populations. Among scaffolds composed of long-lasting porous ceramics and biodegradable natural or synthetic polymers with varying service lives, injectable hydrogels attract attention to regenerate dental pulp due to the capability of filling non-uniform voids such as pulp cavity. In this study, two types of injectable hydrogels were formulized by designing distinct morphologies with the utilization of different biomaterials. Firstly, Tideglusib (Td)-loaded hyaluronic acid hydrogels (HAH) incorporated with Rg1-loaded chitosan microspheres (CSM) were developed for vital pulp regeneration, providing controlled release of Td drug and a ginsenoside Rg1 for matching two basic requirements of pulp; (i) odontoblastic differentiation of human dental pulp stem cells (DPSC) and (ii) vascularization of pulp. Secondly, gelatin methacrylate (GelMA)/thiolated pectin (PecTH) hydrogels (GelMA/PecTH) incorporated with the electrospun core/shell fibers, i.e. melatonin (Mel)-loaded polymethylmethacrylate (PMMA)/Td-loaded silk fibroin (SF) as the core and the shell components respectively, were designed as an alternative injectable hydrogel for vital pulp regeneration, providing controlled release of Td and Mel for inducing proliferation and odontoblastic differentiation of DPSC for a prolonged period. For the first formulation, the expression of the specific genes (COL1A1, ALP, OCN, Axin-2, DSPP, and DMP1) confirmed odontogenic differentiation of DPSC which were incubated with Td-loaded HAH hydrogels incorporated with Rg1-loaded CSM microspheres. Angiogenic potential of the Rg1-containing hydrogel systems among all the other groups were shown on Matrigels in vitro with HUVEC. For the second formulation, cell viability assay showed that DPSC proliferated more in the groups of GelMA/PecTH hydrogels incorporated with PMMA/SF with a Mel/Td ratio of 1/3. ALP activity tests demonstrated that DPSC had the highest odontogenic differentiation level in the hydrogels groups containing PMMA/SF with a Mel/Td ratio of 3/1. Therefore, these novel injectable hydrogels have a potential as candidate biomaterials for vital pulp regeneration.
Deprem yalıtmlı köprülerin salınım davranışının karşılaştırmalı irdelenmesi
In this thesis study a comprehensive roadmap is proposed for detailed modeling of rocking behavior of superstructure deck. Furthermore, a parametric study is conducted to determine effect of the superstructure rocking in enhancing the seismic performance of the box girder type bridge structures. For this purpose, various nonlinear models varying based on one chosen parameter are designed. Nonlinear boundary time history analysis (NTHA) of the models are then conducted being exposed to a set of ground motions scaled with reference to response spectra obtained for a specified coordinate in Canakkale region of Turkey. In the analysis, the effect of different parameters such as number of spans, eccentricity (e) of the bearing lines with respect to pier axis, pier height, span length, friction coefficient of the Friction Pendulum Sliding (FPS) Isolators, radius of curvature of FPS and ground motion scale and intensity are considered. The results of NTHA are then used to discuss the effects of these parameters on the seismic performance of box girder bridges in terms of pier moment and base shear.
Kemik hatalarında dolgu olarak potansiyel kullanım için grafen oksit / lanthanum katkılı kalsiyum fosfat kemik çimentoları
In this thesis, mesoporous particles of β-tricalcium phosphate (βTCP, βCa3(PO4)2) and lanthanum (La) doped βTCP were synthesized by wet precipitation method attached with a microwave refluxing system. Pure dicalcium phosphate (DCP) and La modified dicalcium phosphate (La-DCP) bone cements were prepared based on acid/base reaction between βTCP (or La- βTCP) and monocalcium phosphate monohydrate (MCPM) in the presence of water. DCP bone cements were also mixed with 1.5-3.5wt.% of graphene oxide (GO). The obtained materials were characterized by XRD, FTIR, SEM, ICP-OES, BET, and helium pycnometer. The loading /releasing analysis of proteins on/from βTCP and La- βTCP disks surfaces were evaluated using Fetal bovine serum (FBS) proteins. In vitro cell culture studies were performed with the human osteosarcoma cell line (Saos-2). Upon incorporating La3+ ions, expansion in lattice parameters of βTCP crystal was observed along with inhibition of growth rate. βTCP and La- βTCP materials were mesoporous in nature. Pore diameter and pore volume were expanded with the incorporation of La3+ ions. The outcomes confirmed that all La-βTCP materials were cytocompatible, and strict dose dependent effect of La3+ ions was observed on cell viability and alkaline phosphatase (ALP) activity. In DCP bone cements, the pure phase of brushite transformed into monetite with minimum addition of La3+ ions (0.090 mole), and the plate-like crystals of brushite turned into spheroid particles. The results of in vitro experiments on proliferation, adhesion, and osteogenic differentiation of Saos-2 cells indicated that the addition of 0.225 mole of La3+ ions promoted these properties compared to pure DCP. With the addition of GO to DCP bone cements, the mechanical properties of 2La-DCP cements improved. As the GO/2La-DCP bone cements were biocompatible, proliferation and differentiation properties of cells were significantly improved with the addition of GO.
Giyilebilir kardiyovasküler takip için grafen tekstil tabanlı akıllı giysi
Giyilebilir elektronik, son zamanlarda tüketici elektroniği pazarına başarılı ticari ürünler sunmaya başlayan hızla büyüyen bir alandır. Biyogeribildirim veya kontrol komutları gibi giyilebilir sistemlerde biyopotansiyel sinyallerin istihdam edilmesinin, bakım noktası sağlık izleme sistemleri, rehabilitasyon cihazları, insan-bilgisayar / makine arayüzleri (HCI / HMI'ler) ve beyin-bilgisayar arayüzleri (BCI'ler) dahil olmak üzere birçok teknolojide devrim yaratması beklenmektedir. Elektrotlar bu tür ürünlerin belirleyici bir parçası olarak görüldüğünden, neredeyse on yıldır incelenmiş ve tekstil elektrotlarının ortaya çıkmasına neden olmuştur. Bu çalışma, kişiselleştirilmiş sağlık izleme uygulamaları için giyilebilir elektrokardiyografi (EKG) sensörlerinin geliştirilmesi için grafen nanotekstillerin sentezi ve uygulaması hakkında rapor vermektedir. Bu çalışmada ilk kez elektrokardiyogramın tek bir kola yerleştirilen grafen tekstillerle başarılı bir şekilde elde edildiğini gösterdik. Sadece bir elastik kol bandı ve "tüm tekstil yaklaşımı" kullanımı, kullanıcının kesintisiz kalp izleme sağlamasını kolaylaştırır. Daldırma kaplama ve şablon baskı teknikleri kullanılarak üretilen grafen tekstillerinin işlevselliği, EKG sinyallerinin invaziv olmayan ölçümü, geleneksel ön jelleşmiş, ıslak, gümüş / gümüş-klorür (Ag / AgCl) ile grafen elektrotlatr karşılaştırıldığında % 98'e kadar mükemmel korelasyon ile gösterilmiştir. Farklı durumlarda güçlü bir şekilde elde edilen EKG sinyalleriyle kalp atış hızı oranları başarıyla belirlenmiştir. Burada sunulan sistem düzeyinde entegrasyon ve bütünsel tasarım yaklaşımı, giyilebilir kalp izleme cihazlarında en son teknolojinin geliştirilmesinde etkili olacaktır.
Yapay sinir ağı tabanlı görüntü sınıflandırma tekniği ile X-ray tarama görüntülerinden usb bellek içerenlerin sınıflandırılması
X-ray tarama teknolojisinin gelişmesi ve yaygınlaşması sonucunda, üretilen görüntülerin sayısı da aynı oranda artmaktadır. X-ray tarama sistemleri tıptan endüstriye birçok alanda kullanıldığı gibi güvenlik alanında sıklıkla tercih edilmektedir. Gelişen teknoloji ile yüksek çözünürlüklü görüntü üretme kapasitesine sahip yeni nesil X-ray cihazları güvenlik alanında, potansiyel tehditlerin tespiti konusunda hayati önem taşımaktadır. Ancak kullanım esnasında hızlı tarama yapabilen bu cihazların ürettiği görüntülerinde kullanıcılar tarafından hızlı bir şekilde analiz edilmesi gerekmektedir. Özellikle veri güvenliğini tehdit eden USB Bellek gibi küçük cihazların hızlı bir şekilde Xray görüntülerinden tespiti bir sorun oluşturmaktadır. Bu çalışma, derin öğrenme tabanlı sınıflandırma yöntemleri ile tarama görüntülerinde USB bellek olup olmadığını tespit etmeyi amaçlamaktadır. Bu kapsamda, X-Ray tarama sistemleriyle 1217 adet USB Bellek içeren ve içermeyen veri oluşturulmuştur. Oluşturulan veri seti Evrişimsel Sinir Ağları (ESA) tabanlı derin öğrenme mimarisine sahip 8 farklı model ile eğitilmiştir. Yapılan eğitim sonucunda (%93) başarı oranı ile en yüksek genel doğruluk değerine ResNet50V2 modeli ile ulaşılırken, ResNet50 modeli ile en düşük (%67) genel doğruluk değeri elde edilmiştir. Anahtar Sözcükler: X-ray, Tarama Sistemleri, Yapay Zekâ, Yapay Sinir Ağları, USB bellek, ESA
Doğal halde olmayan proteinlerin istatiksel mekaniği ve yerel dinamiği
Statistical Mechanics and Local Chain Dynamicsof Denaturated ProteinsAbstractBioinformatics is a fast growing research area and describes any use of computers tohandle biological information. One of the major research efforts is structure prediction ofproteins. The torsion angles (phi-psi) of proteins are considered as the degrees of freedomof a protein because of their control of the proteins? three dimensional structures. In thisthesis, we used rotational isomeric state model in order to calculate statistical averagesand correlations for torsional angles of denaturated proteins. For this purpose, wegrouped each consecutive three residues (triplets) starting from first and used moleculardynamics simulations on triplets. Afterwards, we constructed energy maps for the phi-psiangles of the central residue of each triplet considered. Results showed that triplets haveintrinsic propensities for some conformational preferences which favor the choice of thenative state torsional angles and they are context dependent, determined by the aminoacid sequence of the protein. Furthermore, we improved the stochastic weights with theaim of introducing the long range effects in two different approaches: Monte Carlomethod and genetic algorithm method. Besides, we calculated heat capacity as functionof temperatures by statistical mechanics for three different sets of stochastic weightswhich are obtained from molecular dynamics, Monte Carlo method and genetic algorithmmethod. Additionally, we proposed a dynamic rotational isomeric state model analogousto rotational isomeric state model and calculated transition probabilities from one state toanother. The states are chosen as alpha-helix, beta-sheets, turns and all other states.Results support the idea that during folding, secondary structures forms sequentially.
Etan ve metan saflaştırma süreçlerinde kullanılmak üzere metal-organik gözenekli yapılarin gaz ayırma performansının kapsamlı hesaplamalı taraması
Ethane purification and natural gas purification are two energy-intensive processes which can advantage from better separation techniques. Metal organic frameworks (MOFs) are a recent group of nanoporous materials which can be obtained through the combination of metal nodes and organic linkers on different topologies. Considering the large number of available MOFs, it is not possible to fabricate and test the gas separation performance of every single MOF adsorbents and membranes using purely experimental manners. Therefore, the C2H6/C2H4, C2H6/CH4, and CO2/CH4 separation performances of MOFs were investigated using high-throughput computational screening methods in this thesis. In the first part, molecular simulations were used to assess membrane-based C2H6/C2H4 and C2H6/CH4 separation performances of 175 different MOF structures. Results showed that a significant number of MOF membranes is C2H6 selective for C2H6/C2H4 separation in contrast to the traditional nanoporous materials. Several MOFs were identified to exceed the upper bound established for polymeric membranes and many MOF membranes exhibited higher gas permeabilities than zeolites and carbon molecular sieves. In the second part, a multi-level high-throughput computational screening methodology was used to examine the most recent MOF database for membrane-based CO2/CH4 separation. 8 promising MOF membranes offering the best combination of CO2 permeability (>106 Barrer) and CO2/CH4 selectivity (>80) were identified by combining grand canonical Monte Carlo (GCMC) and equilibrium molecular dynamics (EMD) simulations. Permeabilities and selectivities of the mixed matrix membranes (MMM) in which the best MOF candidates were incorporated as filler particles were also investigated. Many MOF membranes could outperform polymeric membranes for CO2/CH4 separation and MOF-based MMMs can have significantly higher CO2 permeabilities and moderately higher selectivities than pure polymers. Computational identification of the promising MOF candidates for CO2 separation depends on the accurate description of electrostatic interactions between CO2 molecules and MOFs. In the last part, role of partial charge assignment methods in high-throughput computational screening of MOFs for CO2/CH4 separation was examined. A quantum based, density-derived electrostatic and chemical charge method (DDEC) and an approximate charge equilibration method (Qeq) were used to compute the adsorption of CO2/CH4 mixture in 1500 MOFs at two different operating conditions. Results showed that the identity of the best performing MOF candidates, which were selected based on the regenerability and adsorbent performance score of MOFs, can change based on the type of the charge assignment method used in simulations. Overall, results showed that high-throughput screening approaches introduced in this thesis can be used to predict gas separation performance of MOF adsorbents and membranes and MOFs can perform better than commercially used materials. The results of this thesis will be useful to guide the experiments to the most promising MOF candidates and to accelerate the development of new MOFs with high performances.
Yapay zeka algoritmalarını kullanarak petrol ve gaz üretim tahminlerinin iyileştirilmesi
The energy industry heavily relies on accurate oil and gas production forecasting, which is crucial in optimizing resource allocation, production planning, and operational efficiency. Predicting future production rates is fundamental to the petroleum industry, enabling businesses to enhance the extraction and refinement of fuels such as gasoline and diesel while minimizing risks and maximizing profits. Reliable forecasting models help stakeholders make informed decisions regarding investment strategies, infrastructure development, and regulatory compliance. As global energy demands continue to rise, the need for precise, data-driven forecasting methodologies becomes increasingly important to ensure stability and sustainability in the petroleum sector. Traditionally, forecasting the future production of energy-rich natural gas and oil has been a widely studied topic in petroleum engineering. Due to their similar extraction processes and market demands, these two resources are often analyzed together. Historically, researchers have relied on conventional techniques such as decline curve analysis and numerical reservoir simulation to estimate future production levels. While these approaches have been widely used, they suffer from several critical limitations. The high computational costs, the extensive time required to complete the simulations, and the dependence on multiple assumptions often result in unreliable and inconsistent predictions. Additionally, these traditional models struggle to adapt to the complex, nonlinear nature of oil and gas production, particularly when dealing with fluctuating reservoir conditions, market dynamics, and external environmental factors. As a result, there is a growing need for more advanced, efficient, and accurate predictive methodologies. In recent years, artificial intelligence (AI) has emerged as a transformative tool in the energy sector, offering the potential to enhance the speed, accuracy, and adaptability of production forecasting. AI-based approaches have significantly reduced the computational time required for forecasting while improving prediction accuracy through automated feature selection, pattern recognition, and model optimization. By leveraging machine learning, ensemble learning, and deep learning techniques, AI-driven models can effectively process large volumes of historical production data to identify trends and make precise forecasts. The integration of AI into petroleum engineering has enabled more efficient decision-making processes, optimized resource utilization, and minimized uncertainties in production planning. This thesis aims to enhance the accuracy and efficiency of oil and gas production forecasting by implementing a robust AI-driven predictive framework. The proposed system employs a combination of machine learning, ensemble learning, and deep learning techniques to develop highly accurate predictive models. By leveraging past production data, the system aims to generate precise future outcome predictions, enabling companies and stakeholders to improve strategic planning, optimize resource allocation, and mitigate market risks. The research focuses on evaluating the performance of multiple AI models to determine the most effective methodology for forecasting oil and gas production with minimal error rates. Eleven different methodologies were used in the proposed system. These consist of five machine learning models: Decision Tree Regressor (DTR), Random Forest Regressor (RFR), K-Nearest Neighbors (KNN), Extreme Gradient Boosting (XGBoost), and Gradient Boosting Regressor (GBR), as well as three ensemble learning models: Bagging, Boosting, and Stacking. Additionally, there are three models in deep learning: Artificial Neural Network (ANN), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM). By incorporating these methodologies, the system ensures that multiple aspects of data-driven forecasting are considered, thereby improving robustness and overall performance. The results of this thesis demonstrate that ensemble learning techniques, particularly the stacking model, offer superior predictive performance compared to standalone machine learning or deep learning models. The ensemble-based approaches significantly reduced error rates in key evaluation metrics, including Mean Absolute Error (MAE) and Mean Squared Error (MSE). Among all the models tested, the stacking model achieved the highest level of accuracy, with an R-squared value of 99%, indicating its exceptional ability to capture complex patterns and dependencies in oil and gas production data. This finding highlights the effectiveness of stacking as a model integration technique, which combines the predictive power of multiple base learners to produce a highly accurate final prediction. Beyond academic contributions, this thesis has significant practical implications for the petroleum industry. The integration of AI-driven forecasting models into oil and gas production workflows can lead to substantial improvements in decision-making efficiency, cost reduction, and risk management. By utilizing intelligent predictive analytics, industry professionals can make data-driven investment decisions, optimize drilling and extraction strategies, and improve overall operational performance. Furthermore, AI-based models provide greater adaptability to changing reservoir conditions and external economic factors, ensuring more reliable and sustainable production forecasts in the long term. The ability to dynamically adjust to real-time production data, external market fluctuations, and environmental constraints makes AI-powered forecasting a critical component of the future energy sector. The adoption of AI-based forecasting methodologies has the potential to revolutionize the petroleum industry, ensuring more precise and reliable production planning in an increasingly complex and dynamic energy market. Additionally, continued advancements in AI and data science will enable the integration of real-time monitoring systems, improving responsiveness to changing market demands and environmental conditions. Generating highly accurate and adaptive predictions will be crucial in meeting global energy needs while promoting efficiency and sustainability in the oil and gas sector. Furthermore, the expansion of AI applications in the energy sector may open avenues for automation, enhanced reservoir management, and predictive maintenance, further optimizing extraction and refining processes. Future research will focus on incorporating additional real-time operational parameters, expanding the dataset to include diverse geological formations, and developing hybrid AI architectures that combine the strengths of multiple methodologies to further enhance forecasting performance. Overall, this thesis provides a comprehensive framework for AI-driven oil and gas production forecasting, demonstrating the immense potential of integrating machine learning, deep learning, and ensemble learning techniques to achieve highly accurate and efficient predictions. The continued evolution of AI methodologies will play a crucial role in shaping the future of energy production, ensuring more sustainable, efficient, and intelligent resource management strategies in the petroleum industry.