Khder I. M., Gheni S. A., Hassan Z. F., Hamd M. I., Türköz Karakullukçu N., Tahah A. K.
Journal of Petroleum Research and Studies , cilt.15, sa.4, ss.181-202, 2025 (Hakemli Dergi)
Özet
The growing demand for solutions to plastic waste and sustainable fuel options
around the world has inspired research into catalytic pyrolysis as a potential
method to convert waste plastic into profitable biofuels. The intensity of pyrolysis
processes af
fected by many process factors makes traditional modeling methods
difficult. This research uses artificial nervous network (ANNS) to create a
prediction model aimed at increasing biofuel conversion in catalyst pyrolysis. The
range of variables considered
in the study, including temperature, residence time,
catalyst type, conversion, density, particular gravity, API, viscosity, and higher
heating value were used to train the ANN model, giving accurate predictions of
biofuels production under various condit
ions. The Levenberg
-
Marquard method
was employed for network training, guaranteed better accuracy and low error. The
comparative comparison of traditional modeling functioning and AI
-
operated
approaches reflect the advantage of the artificial nervous net
work (ANN) model
in real
-
time managing non
-
linear interactions and optimizing processes.
Conclusions suggest that the AI
-
operated approaches clearly promote process
efficiency, reduce waste, and improve decision making in industrial contexts.in
this study
a perfect match was achieved between the predicted data and
experimental data, with R2 value of 1, indicating a perfect alignment between the
predictions and experimental results. This research highlights the ability of
artificial intelligence to increase
permanent chemical engineering functioning and
improve biofue
l production from waste plastic
.The growing demand for solutions to plastic waste and sustainable fuel options
around the world has inspired research into catalytic pyrolysis as a potential
method to convert waste plastic into profitable biofuels. The intensity of pyrolysis
processes af
fected by many process factors makes traditional modeling methods
difficult. This research uses artificial nervous network (ANNS) to create a
prediction model aimed at increasing biofuel conversion in catalyst pyrolysis. The
range of variables considered
in the study, including temperature, residence time,
catalyst type, conversion, density, particular gravity, API, viscosity, and higher
heating value were used to train the ANN model, giving accurate predictions of
biofuels production under various condit
ions. The Levenberg
-
Marquard method
was employed for network training, guaranteed better accuracy and low error. The
comparative comparison of traditional modeling functioning and AI
-
operated
approaches reflect the advantage of the artificial nervous net
work (ANN) model
in real
-
time managing non
-
linear interactions and optimizing processes.
Conclusions suggest that the AI
-
operated approaches clearly promote process
efficiency, reduce waste, and improve decision making in industrial contexts.in
this study
a perfect match was achieved between the predicted data and
experimental data, with R2 value of 1, indicating a perfect alignment between the
predictions and experimental results. This research highlights the ability of
artificial intelligence to increase
permanent chemical engineering functioning and
improve biofue
l production from waste plastic
.