http://lib.buet.ac.bd;localhosthttp://:8080/xmlui/handle/123456789/4 2026-11-13T09:31:48Z http://lib.buet.ac.bd;localhosthttp://:8080/xmlui/handle/123456789/7403 Development of a carbon-based adsorbent from biogenic waste for carbon capture Dr. Kazi Bayzid Kabir; Ummay Salma; 0419024002; 668.62/UMM/2025 Anthropogenic carbon dioxide emission into the atmosphere is over 36 billion tons annually, equivalent to around 1100 metric tons each second. The atmospheric CO2 concentration surpassed 400 parts per million permanently in 2016. Reducing CO2 emissions is the only solution to improve the climatic conditions in the future. A long-term solution in the current scenario is switching to CO2-free energy sources. At the same time, there is a need to cut emissions in the short and medium term through carbon capture. Carbon, as a material, has excellent potential as an adsorbent to remove CO2 from flue gas. The present trend in the ocean of carbon is the functionalization of porous carbon, which is now making history through minor modifications and outstanding improvements worldwide. In addition, production of carbon-based adsorbents from biogenic waste is an alternative to diminish the environmental problems related to agro-industrial waste disposal and the emission of hazardous materials to the environment. Consequently, there exists a plethora of applications of waste derived carbon across diverse sectors. In peruse of this situation, to contribute to CO2 mitigation objectives of this research were standardized. To develop an eco-friendly adsorbent material of higher adsorption capacity using biogenic waste, that performs efficiently in carbon capture applications. To replace traditional carbon sources, for this study wastewater generated from the biomass pretreatment (denoted as PHWW) used as precursor to produce solid carbon through hydrothermal carbonization (HTC). Carbon recovered from the waste utilized as the raw material to develop solid adsorbent for carbon capture. Adsorbent developed for this study used in gas-phase application possess several significant additional characteristics. These include a notable adsorptive capacity per unit volume, a substantial retention capacity, a pronounced preference for adsorbing gases even in the presence of moisture, a minimal resistance to the flow of gases, and a complete release of adsorbates as temperatures increase. Within the design range of HTC, 220°C shows a promising carbonization temperature according to the recovered amount of carbon as hydrochar. In order to get higher surface area and micropore volume activation of HTC char was carried out by both physical and chemical methods with nitrogen flow. For chemical activation KOH was used as activating reagent. After going through several characterizations’ adsorbent produced by chemical activation showed higher surface area 1047.46 m2/g and pore volume 0.6 cm3/g activated at temperature 800°C. However, adsorbents activated at temperature 700°C for 1h showed maximum adsorption capacity 5.34 mmol/g at adsorption temperature 0°C and CO2 partial pressure 0.15 bar. Narrow microporosity and surface area seemed to work together to control this CO2 physisorption. Conversely, surface functionalities were essential for encouraging selective chemisorption of CO2 at higher temperatures and lower partial pressures. Nitrogen functional groups were introduced onto the carbon surface using hydrothermal treatment to boost the adsorption capability at higher temperatures. To see how the nitrogen concentration of the adsorbent changed following treatment, two distinct modifiers- a 25% NH4OH solution and solid CO(NH2)2 were utilized separately. Urea treated adsorbent showed 7.36 mmol/g adsorption capacity at 0°C for partial pressure 0.15 bar which was ~38% greater than the untreated adsorbent. It also provided good performance for cyclic operations. At temperature 80°C, adsorption capacity was found 3.71 mmol/g at partial pressure 0.5 bar of CO2 which was ~488% greater than untreated adsorbent. The adsorbent developed for this research exhibits comparable adsorption capability at critical condition’s application in comparison to previously reported findings. Even for cyclic operation it showed better performance and only 7% capacity was lost after five complete operations. Its superior adsorption capacity, even in humid conditions, makes it a viable option for post-combustion CO2 capture from flue gas as an adsorbent. Compared to conventional processes, there is a huge potential of solid carbon-based adsorbents to remove CO2 from flue gas, especially for regeneration energy and significant cost reduction. Overall, the development of HTC char into CO2 adsorbent from biogenic waste represents a promising step towards a more sustainable and environmentally sustainable future. 2025-05-31T00:00:00Z http://lib.buet.ac.bd;localhosthttp://:8080/xmlui/handle/123456789/7402 Multi-class probabilistic classification approach using Gaussian process and support vector machine Dr. Kais Bin Zaman; Kaniz Fatema; 0422082011; 670.427/KAN/2025 In recent years, Support Vector Machines (SVMs) have demonstrated strong performance in efficiently and accurately handling large-scale classification tasks, often outperforming many existing classification techniques. Meanwhile, Gaussian Process Classification (GPC) offers the distinct advantage of generating probabilistic outputs, providing insights into prediction uncertainty—though this comes at a significant computational cost. This thesis introduces two novel algorithms tailored for multi-class probabilistic classification. These methods integrate the strengths of SVM and GPC into a unified framework, enabling the classification of large datasets in a relatively reduced timeframe while also delivering probabilistic predictions for each class. Such probabilistic outputs are valuable for deriving confidence measures and assessing uncertainty in classification decisions. The effectiveness of the proposed models is evaluated using six diverse classification datasets, ranging from 1,197 to 5,811 data points and 3 to 27 predictors. Each dataset features 3 to 7 categorical outcome classes across health, environmental, and industrial domains. The datasets are sourced from various platforms, such as UCI Machine Learning Repository and Kaggle. One dataset is synthetically generated using Python. Comparative analyses show that these new algorithms achieve substantial computational efficiency, particularly during the prediction phase on unseen data, and often outperform existing state-of-the-art methods. In terms of accuracy, the proposed approaches yield consistently competitive results, sometimes surpassing current leading algorithm’s performance. 2025-07-01T00:00:00Z http://lib.buet.ac.bd;localhosthttp://:8080/xmlui/handle/123456789/7401 Machine learning-based multi-criteria optimization for adaptive hybrid flow shop scheduling Dr. Ferdous Sarwar; Zahin Ar Rafi; 0422082010; 671/ZAH/2025 The Adaptive Hybrid Flow Shop Scheduling Problem (AHFSSP) is a cornerstone of modern manufacturing, involving the coordination of jobs through multiple processing stages, each equipped with parallel machines, under conflicting objectives such as minimizing makespan, tardiness, and energy utilization. As an NP-hard problem, HFSSP poses significant challenges in dynamic industrial environments where uncertainties like fluctuating job arrivals, and varying priorities disrupt traditional scheduling methods. This thesis proposes an innovative hybrid methodology that integrates unsupervised machine learning, evolutionary optimization, and reinforcement learning to address these challenges. Specifically, k-means clustering is employed to group jobs based on critical features, simplifying the scheduling problem by prioritizing high-impact tasks. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) generates a Pareto-optimal set of solutions, effectively balancing the four conflicting objectives to produce diverse, high-quality schedules. Proximal Policy Optimization (PPO), a reinforcement learning algorithm, refines these schedules by shaping rewards and updating policies that respond to real-time changes, such as new job arrivals. The methodology is validated using a real-world dataset from a Fast-Moving Consumer Goods (FMCG) company, comprising 100 jobs across four stages with 12–17 machines per stage. The hybrid model achieved superior performance, with a makespan of 312.9 minutes, tardiness of 94.88 minutes, and a computation time of 0.0203 seconds, outperforming the benchmarks. Statistical validation through ANOVA and T-tests confirmed significant performance improvements, with on-time completion rates of 61%–74% across priority clusters. Comparative analysis against an exact method (Branch and Bound) demonstrated near-optimal solutions with a computation time over 100 times faster, highlighting the model’s suitability for real-time applications. By combining clustering for task prioritization, NSGA-II for global optimization, and PPO for dynamic adaptation, this research delivers a scalable, intelligent, and robust solution for adaptive HFSSP, advancing operational efficiency in complex manufacturing systems. 2025-07-13T00:00:00Z http://lib.buet.ac.bd;localhosthttp://:8080/xmlui/handle/123456789/7400 Life cycle assessment of end-of-life management of multilayer plastic waste Dr. Kazi Bayzid Kabir; Zonaed Sakib; 0423022125; 668.4192/ZON/2025 In recent times, multilayer plastic packaging has been in high demand, as it provides better preservation and a prolonged shelf life for various packaging commodities. How-ever, its complicated structure poses a challenge to the existing recycling system and often ends up in a landfill or is incinerated. Unlike the current management system of MLP, the selective dissolution process represents a paradigm shift, as it enables the re-cycling of MLP with high material recovery and at a reasonable cost. Besides the bene-fits it offers, it is important to assess the environmental footprint of this process to ensure sustainability. Therefore, this study conducts a comparative life cycle assessment be-tween the selective dissolution process and the existing end-of-life management tech-niques (landfilling, incineration) to evaluate whether itbenefits to the existing practices. This study defines 1000 kg post-consumer plastic waste as a functional unit. It assumes 50% single-layer PE is mixed with 50% two-layer MLP (76% PE & 24% PET). This study assesses environmental impact in nine impact categories (fine particulate matter formation, freshwater ecotoxicity, marine ecotoxicity, global warming, human carcino-genic toxicity, human non-carcinogenic toxicity, terrestrial acidification, fossil resource scarcity and water consumption) using the ReCiPe Midpoint (E) impact assessment method. The outcome of this study finds the landfilling of plastic to be the least favora-ble, and the selective dissolution process to be the most favorable in terms of environ-mental sustainability among the alternatives. For example, the freshwater ecotoxicity impact of landfilling is 58.96 kg 1,4-DCB, while for selective dissolution the impact is -64.62 kg 1,4-DCB. Incineration of plastic waste with energy recovery is environmental-ly friendly in all impact assessment categories, except for the global warming impact. The global warming impact of incineration is 297.22 kg CO2 eq. A hotspot analysis of the selective dissolution process identifies that the steam energy required for the reboiler in the distillation column (78%-98%), pretreatment of multilayer film (9%-12%), and the solvent consumed (3%-6%) are the significant contributors to the total impact. Finally, a sensitivity analysis of the selective dissolution process has been performed, varying four key input parameters-5% lower reboiler duty, 95% solvent recovery, natural gas heating, and a 70:30 PE: PET ratio. This analysis shows that switching to natural gas for heating reduces the effect by around 70%-80% for every impact category. 2025-07-16T00:00:00Z