Faculty Publications

Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736

Publications by NITK Faculty

Browse

Author: search.filters.author.Surya, D.V.Subject: search.filters.subject.Machine-learning

Search Results

Now showing 1 - 6 of 6
  • No Thumbnail Available
    Item
    Microwave-assisted In-situ catalytic co-pyrolysis of polypropylene and polystyrene mixtures: Response surface methodology analysis using machine learning
    (Elsevier B.V., 2023) Kamireddi, D.; Terapalli, A.; Sridevi, V.; Tukaram Bai, M.T.; Surya, D.V.; Sankar Rao, C.S.; Jeeru, L.R.
    Polypropylene (PP) and Polystyrene (PS) are the major plastic fractions found in mixed plastic waste. Hence, the current study was focused to convert PP and PS into useful products via microwave-assisted pyrolysis (MAP). In addition, the understanding of feedstock conversion, product yields, and energy requirements in pyrolysis, co-pyrolysis, and catalytic co-pyrolysis was investigated. Experiments were conducted at a constant microwave power of 450 W till the reaction temperature reached up to 600 °C. When PS pyrolyzed, a heating rate of 56 °C/min resulted in 80 wt% of oil yield. Whereas PP pyrolysis produced 42 wt% of oil at a heating rate of 76 °C/min. In the PP: PS co-pyrolysis, the heating rate was decreased to 52 °C/min by yielding 51 wt% of oil. In catalytic co-pyrolysis of PP: PS with KOH resulted in variation in product yields and heating rate. An increase in PS quantity at a constant mass of PP resulted in the enhancement of oil yields from 58 to 84 wt% and a decrease in gas yields. The specific microwave power in the catalytic co-pyrolysis (7–18 W/g) is lower compared to the non-catalytic case (22–30 W/g). Whereas, the pyrolysis time in non-catalytic pyrolysis (7–11 min) is lower compared to catalytic co-pyrolysis (14–37 min). The addition of a catalyst resulted in a decrease (23–50%) in microwave conversion efficiency than that of the non-catalytic case (60–85%). The difference in predicted and actual result analysis proved co-pyrolysis synergy in product formation and energy consumption. © 2023 Elsevier B.V.
  • No Thumbnail Available
    Item
    Utilizing support vector regression modeling to predict pyro product yields from microwave-assisted catalytic co-pyrolysis of biomass and waste plastics
    (Elsevier Ltd, 2023) Ramesh, P.; Sankar Rao, C.S.; Surya, D.V.; Kumar, A.; Basak, T.
    The rise in plastic waste production has led to the development of co-pyrolysis of waste plastics and biomass as a potential solution. This process converts waste into valuable resources, including chemicals and pollutant-absorbing materials. Accurately predicting product yields is crucial and involves considering feedstock characteristics and pyrolysis conditions. No previous work on machine learning (ML) predicts pyro-products considering catalyst and blend as input features. This study used a support vector machine (SVM) to predict pyro-product yields from microwave-assisted co-pyrolysis of biomass and plastics. SVM models were trained, validated, and then applied to new data. The results showed high predictive accuracy, with R2 values of 0.96, 0.93, and 0.91 for bio-oil, biochar, and biogas, respectively. The SVM model demonstrated strong predictive capabilities, indicating effective generalization ability based on statistical parameters. Additionally, SVM models incorporating all features performed better than those based on 'elementary analysis (EA)' and 'proximate analysis (PA)' alone. The pearson correlation coefficient (PCC) approach assessed the correlation between input features to remove highly correlated variables. The partial dependence analysis reveals the individual effects of influential factors and their interactions in the co-pyrolysis process, highlighting significant features like carbon, hydrogen, ash, volatile matter, and nitrogen content that influence oil, char, and gas yields, thereby providing valuable insights for optimization strategies in co-pyrolysis. © 2023 Elsevier Ltd
  • No Thumbnail Available
    Item
    Exploring machine learning applications in chemical production through valorization of biomass, plastics, and petroleum resources: A comprehensive review
    (Elsevier B.V., 2024) Mafat, I.H.; Surya, D.V.; Sharma, S.K.; Sankar Rao, C.
    Machine learning (ML) is a subtype of artificial intelligence that uses a computer's ability to learn from a given set of accessible data. ML is becoming prominent in almost every business, including the domain of chemical engineering, where there have been numerous researches and investigations. This article provides a detailed overview of the use of ML in the production and characterization study of biomass, polymers, and petroleum products. Categories of ML, including classification, regression, and clustering, are also investigated to get a deeper understanding of ML. From this review, it can be concluded that ML has aided in numerous domains, such as the prediction of biomass energy, the stability of crude oil based on NMR spectroscopy, the calculation of gasoline's octane number, the estimation of fuel oil's kinematic viscosity, the classification of waste plastics, and the estimation of drilling efficiency in petroleum reservoirs, among others. Apart from this, ML has also been playing a significant role in the microwave-assisted pyrolysis of biomass, polymers, and petroleum resources. ML substantially influences chemical engineering and is especially useful for enhancing system efficiency and monitoring processes that are difficult to understand manually. Although several obstacles are associated with ML, such as black box behavior, the need for a large amount of data, and the difficulty of understanding the predictions, deploying the model in the future is uncomplicated once the learning program has been trained. © 2024 Elsevier B.V.
  • No Thumbnail Available
    Item
    A review on the role of various machine learning algorithms in microwave-assisted pyrolysis of lignocellulosic biomass waste
    (Academic Press, 2024) Mafat, I.H.; Surya, D.V.; Sankar Rao, C.S.; Kandya, A.; Basak, T.
    The fourth industrial revolution will heavily rely on machine learning (ML). The rationale is that these strategies make various business operations in many sectors easier. ML modeling is the discovery of hidden patterns between multiple process parameters and accurately predicting the test values. ML has provided a wide range of applications in Chemical Engineering. One major application of ML can be found in the microwave-assisted pyrolysis (MAP) of lignocellulose bio-waste. MAP is an energy-efficient technology to obtain high-saturated hydrogen-rich liquid fuels. The main focus of this review study is understanding the utilization of various types of ML algorithms, including supervised and unsupervised techniques in microwave-assisted heating techniques for diverse biomass feedstocks, including waste materials like used tea powder, wood blocks, kraft lignin, and others. In addition to developing effective ML-based models, alternative traditional modeling approaches are also explored. In addition to various thermochemical conversion processes for biomass, MAP is also briefly reviewed with several case studies from the literature. The conventional modeling methodology for biomass pyrolysis with microwave heating is also discussed for comparison with ML-based modeling methodologies. © 2024 Elsevier Ltd
  • No Thumbnail Available
    Item
    Development of machine learning model for the prediction of selectivity to light olefins from catalytic cracking of hydrocarbons
    (Elsevier Ltd, 2025) Mafat, I.H.; Sharma, S.K.; Surya, D.V.; Sankar Rao, C.S.; Maity, U.; Barupal, A.; Jasra, R.
    Light olefins are the primary building block for the production of petrochemicals and polymers. Light olefins are largely produced from steam/catalytic cracking of naphtha or ethane/propane. Selectivity to light olefins is significantly dependent on the reaction conditions. In this article, several machine learning models are developed and tested to predict the selectivity of ethylene and propylene using seven input features. For this study, a total of eight ML models consisting of adaptive boost, extreme gradient boost, categorical boost, light gradient boost, decision tree with bagging, random forest, k-nearest neighbour, and artificial neural models are developed. The extreme gradient boost model gave the highest prediction accuracy for the ethylene selectivity, while the light gradient boost gave the highest R2 for the propylene selectivity. The SHAP analysis showed the input parameter's importance ranking for ethylene predictions as temperature > number of carbon atoms > Si/Al ratio > acidity > weight hourly space velocity > effect of diluent > number of hydrogen atoms. The importance ranking of input parameters for propylene selectivity was observed as weight hourly space velocity > acidity > temperature > Si/Al ratio > effect of diluent > number of carbon atoms > number of hydrogen atoms. © 2024 Elsevier Ltd
  • No Thumbnail Available
    Item
    Hospital plastic waste valorization through microwave-assisted Pyrolysis: Experimental and modeling studies via machine learning
    (Elsevier Ltd, 2025) Ramesh, R.; Sankar Rao, C.; Surya, D.V.; Kumar, A.
    The COVID-19 pandemic generated a global upsurge in hospital plastic waste (HPW) as a consequence of the widespread utilization of personal protective equipment (PPE) composed of diverse polymer materials. The constant demand for PPE worldwide led to the accumulation of substantial volumes of high-polymer-based plastic waste. To tackle this challenge, researchers delved into the conversion of HPW into valuable chemicals through a process known as microwave-assisted pyrolysis (MAP). This method entails the transformation of HPW into high-quality char and liquid oil, which can serve as a source of fuel. In this study, our primary focus was to understand how the ratio of HPW (hospital plastic waste) to susceptor weight influenced the yields and characteristics of the resulting products in the context of the MAP process. To facilitate the experimental setup, a Central Composite Design (CCD) was employed. The impact of varying HPW weights and susceptor quantities on the production of value-added products was investigated. The analysis of condensed organic vapor decomposition revealed an increase in liquid yields (73.6 wt %, 76.6 wt %, 80.7 wt %) as the graphite content increased at a constant 30 g HPW. Conversely, gas yield decreased with higher susceptor and HPW quantity. Keeping the graphite constant at 4g, the gas yield declined (32.5 wt %, 30.7 wt %, and 24.7 wt %) as HPW increased. Additionally, gas yield exhibited a drop (32.5 wt % to 18.1 wt %) with an increase in both graphite and HPW. Furthermore, the residual yield decreased (from 1.7 wt % to 1.2 wt %) with a 30 g increase in HPW. In-depth analysis incorporated machine learning techniques to understand the behavior of response variables about susceptor and HPW quantities. The optimization of the MAP process for HPW encompassed various supplementary operational parameters, including susceptor thermal energy, average heating rate, microwave energy, specific microwave power, and product yields. Moreover, the residue generated from the MAP of HPW underwent characterization through X-ray diffraction (XRD), FTIR, and BET analysis. © 2025 Elsevier Ltd