Transformation rule-based molecular evolution for automatic gasoline molecule design

Cai, Guangqing; Liu, Zhefu; Zhang, Linzhou

doi:10.1016/j.ces.2022.118119

Cited by 3 publications

(2 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The characterization data needed to be processed into computer-recognizable and readable digital information. The main methods for the digital representation of petroleum molecules are structure-oriented lumping (SOL), − bond-electron matrix (BEM), , molecular type and homologous series (MTHS) matrix, − structural unit and bond-electron matrix hybrid framework (SU-BEM), and chemoinformatic language such as SMILES. , Each of these representations has its own advantages and disadvantages. Researchers have developed a large number of molecular-level compositional and reaction kinetic models based on digital representations of these petroleum molecules.…”

Section: Vgo Molecular Databasementioning

confidence: 99%

Compositional Modeling of VGO Fractions via Analytical Characterization, Structural Assignment, Property Modeling, and Molecular Component Parsing

Jiang,

Cai,

et al. 2023

Energy Fuels

Self Cite

View full text Add to dashboard Cite

The detailed molecular characterization of the vacuum gas oil (VGO) fractions is the key to understanding petroleum structure−property relationships and optimizing petroleum refinery operations. In our study, a rapid VGO database-based molecular component parsing method was proposed that could directly calculate detailed molecular composition information using conventional bulk properties as inputs. The molecular database of VGO included four aspects: analytical characterization, molecular structural assignment, single molecular property estimation, and bulk property prediction. According to our established molecular database of typical VGO samples, two main parts in the rapid parsing algorithm of the molecular components were present. The method screened the VGO samples with the most similar bulk properties. The molecular component contents of the most similar sample were subsequently optimized by a simulated annealing algorithm, thereby finally obtaining the practical molecular composition of VGO. Our proposed method overcomes the drawbacks of the virtual molecular information acquired by the traditional compositional reconstruction algorithm. Our results were more accurate, obtained faster, and could facilitate a solid foundation for refineries to achieve the optimization of the entire molecular process simulation.

show abstract

Section: Vgo Molecular Databasementioning

confidence: 99%

Compositional Modeling of VGO Fractions via Analytical Characterization, Structural Assignment, Property Modeling, and Molecular Component Parsing

Jiang,

Cai,

et al. 2023

Energy Fuels

Self Cite

View full text Add to dashboard Cite

show abstract

“…In fuel design, a fuel is defined by its physicochemical properties, while its molecular structure or composition is considered a degree of freedom. Potential fuel molecules can be compiled from database screenings − or generated using computer-aided molecular design. ,− The properties can be either obtained from databases or predictive models ,,,, or from experiments. , …”

Section: Introductionmentioning

confidence: 99%

Fuel Ignition Delay Maps for Molecularly Controlled Combustion

Neumann,

Rittig,

Letaief

et al. 2024

Energy Fuels

View full text Add to dashboard Cite

Molecularly controlled combustion systems (MCCSs) combine the advantages of compression-ignition and sparkignition engines by employing both a low reactivity fuel and a high reactivity fuel (HRF). To optimize the MCCS, fuels must be tailored to the engine requirements with respect to fuel reactivity. We aim at deriving requirements for HRFs and identification of suitable HRF candidates. Single-cylinder experiments are performed to assess the suitability of conventional reactivity indicators for MCCS. Fuel ignition delay time (IDT) maps are proposed as alternative reactivity indicators conveying more information. The maps are constructed through five representative IDT measurements at varying temperatures and pressures within the constant volume combustion chamber of the advanced fuel ignition delay analyzer (AFIDA). Specifically, the IDT maps cover the temperature and pressure range of 500−700 °C and 10−35 bar, respectively; and the IDT is measured at the four corner points of that range and the center point. We present measured IDT maps for more than 50 oxygenated and nonoxygenated hydrocarbon species and analyze suitability for MCCSs, e.g., 1,3-dioxane and 1-heptanol are found as possible HRF. The measurement data are further utilized to develop a graph neural network (GNN) model that can predict IDT maps directly from the molecular structure with high accuracy, constituting a first step toward in-silico screening of HRFs for MCCSs.

show abstract

Artificial intelligence for novel fuel design

Sarathy,

Eraqi

2024

Proceedings of the Combustion Institute

View full text Add to dashboard Cite

Transformation rule-based molecular evolution for automatic gasoline molecule design

Cited by 3 publications

References 63 publications

Compositional Modeling of VGO Fractions via Analytical Characterization, Structural Assignment, Property Modeling, and Molecular Component Parsing

Compositional Modeling of VGO Fractions via Analytical Characterization, Structural Assignment, Property Modeling, and Molecular Component Parsing

Fuel Ignition Delay Maps for Molecularly Controlled Combustion

Artificial intelligence for novel fuel design

Contact Info

Product

Resources

About