Investigation of the Effect of Al2O3 - TiO2 Nano Particles on Engine Noise Using Machine Learning Method DOI
Mehmet Selman GÖKMEN, Hasan Aydoğan

Published: July 21, 2023

In this study, the effects of different ratios Al 2 O xmlns:xlink="http://www.w3.org/1999/xlink">3 and TiO nano particles on exhaust noise were analyzed using statistical machine learning methods, employing a 1.2 TSI engine. Response surface methodology was used for experimental design, where ratios, as well engine speed input parameters, determined factors, values selected output parameters.

Language: Английский

Prediction of combustion, performance, and emission parameters of ethanol powered spark ignition engine using ensemble Least Squares boosting machine learning algorithms DOI

D. Jesu Godwin,

Edwin Geo Varuvel, Leenus Jesu Martin

et al.

Journal of Cleaner Production, Journal Year: 2023, Volume and Issue: 421, P. 138401 - 138401

Published: Aug. 9, 2023

Language: Английский

Citations

32
Chemical SuperLearner (ChemSL) - An automated machine learning framework for building physical and chemical properties model DOI
Balaji Mohan,

Junseok Chang

Chemical Engineering Science, Journal Year: 2024, Volume and Issue: 294, P. 120111 - 120111

Published: April 16, 2024

Language: Английский

Citations

5
A Case Study on Integrating an AI System into the Fuel Blending Process in a Chemical Refinery DOI Creative Commons
Abdul Gani Abdul Jameel

ChemEngineering, Journal Year: 2025, Volume and Issue: 9(1), P. 4 - 4

Published: Jan. 3, 2025

Fuel blending plays a very important role in petroleum refineries, because it directly affects the quality of end products, as well overall profitability refinery. This process involves combination various hydrocarbon streams to make fuels that meet specific performance standards and comply with regulatory guidelines. For many decades, most refineries have been dependent on linear programming (LP) models for developing recipes optimization. However, LP normally fail capture complex nonlinear interaction blend components fuel properties, leading off-specification products may necessitate re-blending. work discusses case study hybrid artificial intelligence (AI)-based method gasoline based genetic algorithm (GA) combined an neural network (ANN). AI-based systems are more flexible will enable product specifications regularly result cost reduction owing fall giveaways. The AI-powered discussed can predict, much better accuracy, critical combustion properties such Research Octane Number (RON), Motor (MON), Antiknock Index (AKI), compared classical models, added advantage optimization ratio real time. results showed AI-integrated system was able produce mean absolute error (MAE) 1.4 AKI. obtained MAE is close experimental uncertainty 0.5 octane. A high coefficient determination (R2) 0.99 also when validated new set 57 comprising primary reference blends. highlights potential transforming traditional practices towards sustainable economically viable refinery operations.

Language: Английский

Citations

0
Advancing Sustainable Aviation Fuel Design: Machine Learning for High-Energy-Density Liquid Polycyclic Hydrocarbons DOI
Dilip Rijal,

Vladislav Vasilyev,

Feng Wang

et al.

Energy & Fuels, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 30, 2025

This study explores the identification of polycyclic hydrocarbons (PCHCs) with high energy density (HED) using machine learning (ML) techniques, specifically focusing on establishing a quantitative structure–property relationship (QSPR). The support vector (SVM) algorithm was employed for its strong predictive performance net heat combustion (NHOC), achieving coefficient determination (R2) and low mean absolute error (MAE) 27.821 kJ/mol 20% test data only six key descriptors. From reputable scientific literature databases, 35 potential HED PCHCs (ranging from C6 to C15) were identified. Structural analysis showed that these predominantly consist saturated alkanes featuring multiple triangular, rectangular, pentagonal rings, highlighting significant role strain in HED. emphasizes importance specific as primary considerations sustainable aviation fuel (SAF) design, while also recognizing need meet additional properties comply ASTM D7566/D4054 standards. work successfully achieves initial objectives our SAF program, laying robust foundation further development high-performance, fuels.

Language: Английский

Citations

0
Comprehensive accurate prediction of critical jet fuel properties with multiple machine learning models DOI

Yitong Shao,

Mengxian Yu,

Mengchao Zhao

et al.

Chemical Engineering Science, Journal Year: 2024, Volume and Issue: unknown, P. 121018 - 121018

Published: Nov. 1, 2024

Language: Английский

Citations

3
Artificial intelligence for novel fuel design DOI
S. Mani Sarathy, Basem A. Eraqi

Proceedings of the Combustion Institute, Journal Year: 2024, Volume and Issue: 40(1-4), P. 105630 - 105630

Published: Jan. 1, 2024

Language: Английский

Citations

2
Advancing CO2 Solubility Prediction in Brine Solutions with Explainable Artificial Intelligence for Sustainable Subsurface Storage DOI Open Access
Amin Shokrollahi, Afshin Tatar, Abbas Zeinijahromi

et al.

Sustainability, Journal Year: 2024, Volume and Issue: 16(17), P. 7273 - 7273

Published: Aug. 23, 2024

Underground CO2 storage is crucial for sustainability as it reduces greenhouse gas (GHG) emissions, helping mitigate climate change and protect the environment. This research explores use of Explainable Artificial Intelligence (XAI) to enhance predictive modelling solubility in brine solutions. Employing Random Forest (RF) models, study integrates Shapley Additive exPlanations (SHAP) analysis uncover complex relationships between key variables, including pressure (P), temperature (T), salinity, ionic composition. Our findings indicate that while P T are primary factors, contributions salinity specific ions, notably chloride ions (Cl−), essential accurate predictions. The RF model exhibited high accuracy, precision, stability, effectively predicting even brines not included during training evidenced by R2 values greater than 0.96 validation testing samples. Additionally, stability assessment showed Root Mean Squared Error (RMSE) spans 8.4 9.0 100 different randomness, which shows good stability. SHAP provided valuable insights into feature interactions, revealing dependencies, particularly strength. These offer practical guidelines optimising mitigating associated risks. By improving accuracy transparency predictions, this supports more effective sustainable strategies, contributing overall goal reducing emissions combating change.

Language: Английский

Citations

2
Molecular Profile, Fuel Properties, Engine Performance and Emission Characteristics of Gasoline-Like Fuel Produced Via Cracking of Used Engine Oil Using Na-Fe3O4/HZSM-5 Catalyst DOI Creative Commons

Yaquba M. Sahabi,

Abdullahi Muhammad Sokoto,

Misbahu Ladan Mohammed

et al.

Caliphate Journal of Science and Technology, Journal Year: 2024, Volume and Issue: 6(1), P. 93 - 102

Published: April 11, 2024

A vast quantity of used engine oil (UEO) is generated every day and poses a major disposal issue in modern society due to the heavy metals other hazardous contaminants present it. Due its high carbon content, UEO has great potential be utilized as feedstock for fuel production. The studies on molecular profile, properties, characteristics gasoline-like produced via cracking were conducted. Fe3O4 nanoparticles was synthesized one-spot method using iron (III) chloride hexahydrate (FeCl3.6H2O) (II) tetrahydrate (FeCl2.4H2O) precursors, while HZSM-5 Al2(SO4)3.18H2O Na2SiO3 sources alumina silica, respectively. cracked fixed stainless-steel batch reactor 1h at varying temperature (350 – 450 ⁰C). liquid product obtained analysed composition GC-MS FTIR, ASTM standard procedure determine properties. results showed that catalyst 97.60% selective gasoline range hydrocarbons, which could attributed surface area HZSM-5, offers more active sites catalytic cracking. properties determined include specific gravity (0.76), kinematic viscosity (1.69 mm2 /Sec), flash point (-42°C), auto ignition (225°C), residue (0.12%), lower heating value (40,443 KJ/kg), octane number (94). comparable those commercially available gasoline. Based obtained, it concluded directly spark-ignition engines without any negative impact performance.

Language: Английский

Citations

1
Machine learning of weighted superposition attraction algorithm for optimization diesel engine performance and emission fueled with butanol-diesel biofuel DOI Creative Commons
Ibham Veza, Aslan Deniz Karaoğlan, Şener Akpınar

et al.

Ain Shams Engineering Journal, Journal Year: 2024, Volume and Issue: unknown, P. 103126 - 103126

Published: Oct. 1, 2024

Language: Английский

Citations

1
Fuel type recognition of ester isomer additives in flames by optical diagnostics coupled with machine learning method DOI
J.-W. He, MengFei Chen,

Bingkun Wu

et al.

Science China Technological Sciences, Journal Year: 2024, Volume and Issue: 67(11), P. 3431 - 3442

Published: Oct. 22, 2024

Language: Английский

Citations

1