Abstract
Flashpoint of organic materials is a crucial physical property in industrial applications and laboratory experiments, which provides information on safety standards and needed precautions in handling various organic materials. Proposed methods to determine the flashpoint of an organic material suffer from dependency on other physical properties of the chemical or demand complicated calculations which are time-consuming. In this work, a direct system model is proposed to anticipate the flashpoints of organic materials for a wide range of chemical compounds. The following models of genetic algorithm-adaptive neuro-fuzzy inference system, the least-squares version of the support vector machine, particle swarm optimization-adaptive neuro-fuzzy inference system, and artificial neural network were applied to develop the model. This system model can speed up the flashpoint determination process and its accuracy. It can also anticipate the flashpoints of new organic materials. 79 functional groups were gathered to form a group contribution method to estimate the flashpoints of various chemical compounds. The functional groups were correlated with the model, known as the committee machine intelligent system, which performs accurately to prognosticate the flashpoint for every compound. 1,378 chemical compounds of different chemical categories were used to develop the model, making it suitable to estimate the flashpoints. This system model can determine or anticipate the flashpoints of organic materials with high accuracy and provide useful information for safety considerations in laboratory and industrial applications.
Recommended Citation
F. Amirkhani et al., "Estimating Flashpoints of Fuels and Chemical Compounds using Hybrid Machine-Learning Techniques," Fuel, vol. 323, article no. 124292, Elsevier, Sep 2022.
The definitive version is available at https://doi.org/10.1016/j.fuel.2022.124292
Department(s)
Chemical and Biochemical Engineering
Keywords and Phrases
CMIS; Flashpoint; Group contribution (GC) method; Modelling; Prediction
International Standard Serial Number (ISSN)
0016-2361
Document Type
Article - Journal
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2024 Elsevier, All rights reserved.
Publication Date
01 Sep 2022
Comments
European Commission, Grant CZ.02.1.01/0.0/0.0/15_003/0000456