A5013089411- Machine Learning in Materials Science
- Computational Drug Discovery Methods
- Fault Detection and Control Systems
- Water Quality Monitoring and Analysis
- Advanced Data Processing Techniques
- Polymer crystallization and properties
- Process Optimization and Integration
- Innovative Microfluidic and Catalytic Techniques Innovation
- Reservoir Engineering and Simulation Methods
- Chemical Thermodynamics and Molecular Structure
- Chemistry and Chemical Engineering
- Protein Structure and Dynamics
- Nuclear Engineering Thermal-Hydraulics
- Thermal and Kinetic Analysis
- Catalysts for Methane Reforming
- Thermochemical Biomass Conversion Processes
- Catalytic Processes in Materials Science
- Pharmaceutical Economics and Policy
- Polymer Nanocomposites and Properties
- Heat transfer and supercritical fluids
- Machine Learning and Algorithms
- Nuclear reactor physics and engineering
- Chemical Synthesis and Analysis
- Toxic Organic Pollutants Impact
- Scientific Computing and Data Management
Ghent University
2020-2024
Ghent University Hospital
2022-2023
Chemical engineers rely on models for design, research, and daily decision-making, often with potentially large financial safety implications. Previous efforts a few decades ago to combine artificial intelligence chemical engineering modeling were unable fulfill the expectations. In last five years, increasing availability of data computational resources has led resurgence in machine learning-based research. Many recent have facilitated roll-out learning techniques research field by...
Machine learning has proven effective for predicting properties of pure compounds from molecular structures, but mixtures, in particular oil fractions, are rarely dealt with. At best, the bulk estimated based on compound properties, linear mixing rules, and a reconstructed composition feedstock. As detailed such mixtures is well determined often approximated by lumps, accuracy can be improved. In this work, we demonstrate naphtha case study our property estimation method. First, PIONA...
By combining machine learning with the design of experiments, thereby achieving so-called active learning, more efficient and cheaper research can be conducted. Machine algorithms are flexible better than traditional experiment at investigating processes spanning all length scales chemical engineering. While maturing, their applications falling behind. In this article, three types challenges presented by learning—namely, convincing experimental researcher, flexibility data creation,...
Abstract The challenge of devising pathways for organic synthesis remains a central issue in the field medicinal chemistry. Over span six decades, computer-aided planning has given rise to plethora potent tools formulating synthetic routes. Nevertheless, significant expert task still looms: determining appropriate solvent, catalyst, and reagents when provided with set reactants achieve optimize desired product specific step process. Typically, chemists identify key functional groups rings...
Chemical engineers heavily rely on precise knowledge of physicochemical properties to model chemical processes. Despite the growing popularity deep learning, it is only rarely applied for property prediction due data scarcity and limited accuracy compounds in industrially-relevant areas space. Herein, we present a geometric learning framework predicting gas- liquid-phase based novel quantum datasets comprising 124,000 molecules. Our findings reveal that necessity quantum-chemical information...
Computer-aided synthesis has received much attention in recent years. It is a challenging topic itself, due to the high dimensionality of chemical and reaction space. becomes even more when aim suggest syntheses that can be performed continuous flow. Though flow offers many potential benefits, not all reactions are suited operated continuously. In this work, three machine learning models have been developed provide an assessment whether given may benefit from operation, what likelihood...
Accurate thermochemistry estimation of polycyclic molecules is crucial for kinetic modeling chemical processes that use renewable and alternative feedstocks. In model generators, molecular properties are estimated rapidly with group additivity, but this method known to have limitations structures. This issue has been resolved in our work by combining a geometry-based representation deep neural network trained on ab initio data. Each molecule transformed into probabilistic vector from its...
The effect of catalyst synthesis and reaction conditions on catalytic activity were accurately predicted with an interpretable data-driven strategy. method is demonstrated for CO 2 methanation extendable to other processes.
Abstract Detailed kinetic models play a crucial role in comprehending and enhancing chemical processes. A cornerstone of these is accurate thermodynamic properties, ensuring fundamental insights into the processes they describe. The prediction thermochemical properties presents an opportunity for machine learning, given challenges associated with their experimental or quantum determination. This study reviews recent advancements predicting gas-phase, liquid-phase, catalytic within modeling....
The current kinetic models for polystyrene (PS) pyrolysis contain many simplifications to reduce their size and the corresponding simulation time. Moreover, they are often based on rate coefficients determined using non-ideal experimental data featuring ambiguous process conditions with respect mixing temperature uniformity. In present work, a lab-scale tree-based Monte Carlo ( k MC) model is presented that differentiates between 18 reaction families 26 end-group pairs study product yield...
By combining machine learning with design of experiments, so-called active learning, more efficient and cheaper research can be conducted. Machine algorithms are flexible, better at investigating the processes spanning all length scales chemical engineering. While maturing, its applications lacking behind. Three types challenges faced by identified ways to overcome them discussed: convincing experimental researcher, flexibility data creation, robustness algorithms. A bright future lies ahead...