XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 859
up new avenues for innovation in mineral processing and
related industrial applications.
Future Work
Looking ahead, the study outlines several promising ave-
nues for further exploration and practical application:
• Implementation in Industrial Scale Plant: The
immediate next step involves deploying the devel-
oped algorithm in a full-scale industrial plant. This
crucial phase will validate the algorithm’s practical
efficacy and reliability in a real-world setting, provid-
ing valuable insights for further refinement.
• Transfer Learning: Investigating transfer learning to
enhance data efficiency and model robustness. This
approach facilitates rapid deployment of intelligent
control systems across different plant setups, adapt-
ing previously learned knowledge to new environ-
ments with minimal need for retraining.
• Grey-Box Models: Exploring the integration of
grey-box models, which combine physical laws with
data-driven insights, offers a more flexible and effi-
cient approach to modeling of digital twins. Such
models can provide a deeper understanding of the
underlying processes while still leveraging the power
of machine learning.
• Hybrid Implementation Strategy: Developing a
hybrid approach that merges cloud-based learning
with localized edge computing. This strategy aims
to enhance inter-plant collaboration and provide tai-
lored solutions to each plant’s unique requirements,
balancing centralized insights with local operational
needs.
In summary, this research not only marks a significant
advancement in the field of intelligent manufacturing but
also sets a solid foundation for the future of mineral pro-
cessing. The continued exploration and application of these
innovative strategies promise to bring about substantial
improvements in efficiency, sustainability, and adaptability
within the industry. Emphasizing environmental impact,
this approach has the potential to significantly reduce
energy consumption and costs, contributing to more sus-
tainable practices in mineral processing. The journey into
these new territories is expected to yield novel solutions and
deeper insights into complex industrial systems, further
driving progress, and operational efficiency in the mineral
processing sector.
REFERENCES
Conradie, A.V., &Aldrich, C. (2001, October).
Neurocontrol of a ball mill grinding circuit using
evolutionary reinforcement learning. Minerals
Engineering, 14, 1277–1294. doi:10.1016
/s0892-6875(01)00144-3.
Costea, C.R., Silaghi, H., Gergely, E.I., Husi, G., Coroiu,
L., &Nagy, Z. (2014, November). Approach of PID
controller tuning for ball mill. 2014 International
Symposium on Fundamentals of Electrical Engineering
(ISFEE). IEEE. doi:10.1109/isfee.2014.7050606.
Cronrath, C., Aderiani, A.R., &Lennartson, B.
(2019, August). Enhancing Digital Twins through
Reinforcement Learning. 2019 IEEE 15th International
Conference on Automation Science and Engineering
(CASE). IEEE. doi:10.1109/coase.2019.8842888.
Edwards, R., Vien, A., &Perry, R. (2002). Strategies for
the instrumentation and control of grinding circuits. In
A.L. Mular, D.N. Halbe, &D.J. Barratt (Eds.), Mineral
Processing Plant Design, Practice, and Control. Society for
Mining Metallurgy and Exploration (pp. 2130–2151).
Forssberg, E., &Yanmin, W. (2003). International overview
and outlook on comminution technology. Retrieved
from https://www.semanticscholar.org/paper/19f7ec5
489308b20881de1fbbf07b078a105f5d1.
García, J., &Fernández, F. (2015). A comprehen-
sive survey on safe reinforcement learning.
doi:10.5555/2789272.2886795.
Hochreiter, S., &Schmidhuber, J. (1997, November).
Long Short-Term Memory. Neural Computation, 9,
1735–1780. doi:10.1162/neco.1997.9.8.1735.
Hosten, C., &Fidan, B. (2012, May). An industrial com-
parative study of cement clinker grinding systems
regarding the specific energy consumption and cement
properties. Powder Technology, pp. 183–188. doi:
doi:10.1016/j.powtec.2011.12.065.
Ivezić, D.D., &Petrović, T.B. (2003, June). New approach
to milling circuit control—robust inverse Nyquist array
design. International Journal of Mineral Processing, 70,
171–182. doi:10.1016/s0301-7516(03)00022-x.
Leonel, J. (2018). Perceptrons. Perceptrons. Retrieved
from https://medium.com/@jorgesleonel/perceptrons
-a22fb29facc4.
Pomerleau, A., Hodouin, D., Desbiens, A., &Gagnon, É.
(2000, March). A survey of grinding circuit control
methods: from decentralized PID controllers to multi-
variable predictive controllers. Powder Technology, 108,
103–115. doi:10.1016/s0032-5910(99)00207-7.
up new avenues for innovation in mineral processing and
related industrial applications.
Future Work
Looking ahead, the study outlines several promising ave-
nues for further exploration and practical application:
• Implementation in Industrial Scale Plant: The
immediate next step involves deploying the devel-
oped algorithm in a full-scale industrial plant. This
crucial phase will validate the algorithm’s practical
efficacy and reliability in a real-world setting, provid-
ing valuable insights for further refinement.
• Transfer Learning: Investigating transfer learning to
enhance data efficiency and model robustness. This
approach facilitates rapid deployment of intelligent
control systems across different plant setups, adapt-
ing previously learned knowledge to new environ-
ments with minimal need for retraining.
• Grey-Box Models: Exploring the integration of
grey-box models, which combine physical laws with
data-driven insights, offers a more flexible and effi-
cient approach to modeling of digital twins. Such
models can provide a deeper understanding of the
underlying processes while still leveraging the power
of machine learning.
• Hybrid Implementation Strategy: Developing a
hybrid approach that merges cloud-based learning
with localized edge computing. This strategy aims
to enhance inter-plant collaboration and provide tai-
lored solutions to each plant’s unique requirements,
balancing centralized insights with local operational
needs.
In summary, this research not only marks a significant
advancement in the field of intelligent manufacturing but
also sets a solid foundation for the future of mineral pro-
cessing. The continued exploration and application of these
innovative strategies promise to bring about substantial
improvements in efficiency, sustainability, and adaptability
within the industry. Emphasizing environmental impact,
this approach has the potential to significantly reduce
energy consumption and costs, contributing to more sus-
tainable practices in mineral processing. The journey into
these new territories is expected to yield novel solutions and
deeper insights into complex industrial systems, further
driving progress, and operational efficiency in the mineral
processing sector.
REFERENCES
Conradie, A.V., &Aldrich, C. (2001, October).
Neurocontrol of a ball mill grinding circuit using
evolutionary reinforcement learning. Minerals
Engineering, 14, 1277–1294. doi:10.1016
/s0892-6875(01)00144-3.
Costea, C.R., Silaghi, H., Gergely, E.I., Husi, G., Coroiu,
L., &Nagy, Z. (2014, November). Approach of PID
controller tuning for ball mill. 2014 International
Symposium on Fundamentals of Electrical Engineering
(ISFEE). IEEE. doi:10.1109/isfee.2014.7050606.
Cronrath, C., Aderiani, A.R., &Lennartson, B.
(2019, August). Enhancing Digital Twins through
Reinforcement Learning. 2019 IEEE 15th International
Conference on Automation Science and Engineering
(CASE). IEEE. doi:10.1109/coase.2019.8842888.
Edwards, R., Vien, A., &Perry, R. (2002). Strategies for
the instrumentation and control of grinding circuits. In
A.L. Mular, D.N. Halbe, &D.J. Barratt (Eds.), Mineral
Processing Plant Design, Practice, and Control. Society for
Mining Metallurgy and Exploration (pp. 2130–2151).
Forssberg, E., &Yanmin, W. (2003). International overview
and outlook on comminution technology. Retrieved
from https://www.semanticscholar.org/paper/19f7ec5
489308b20881de1fbbf07b078a105f5d1.
García, J., &Fernández, F. (2015). A comprehen-
sive survey on safe reinforcement learning.
doi:10.5555/2789272.2886795.
Hochreiter, S., &Schmidhuber, J. (1997, November).
Long Short-Term Memory. Neural Computation, 9,
1735–1780. doi:10.1162/neco.1997.9.8.1735.
Hosten, C., &Fidan, B. (2012, May). An industrial com-
parative study of cement clinker grinding systems
regarding the specific energy consumption and cement
properties. Powder Technology, pp. 183–188. doi:
doi:10.1016/j.powtec.2011.12.065.
Ivezić, D.D., &Petrović, T.B. (2003, June). New approach
to milling circuit control—robust inverse Nyquist array
design. International Journal of Mineral Processing, 70,
171–182. doi:10.1016/s0301-7516(03)00022-x.
Leonel, J. (2018). Perceptrons. Perceptrons. Retrieved
from https://medium.com/@jorgesleonel/perceptrons
-a22fb29facc4.
Pomerleau, A., Hodouin, D., Desbiens, A., &Gagnon, É.
(2000, March). A survey of grinding circuit control
methods: from decentralized PID controllers to multi-
variable predictive controllers. Powder Technology, 108,
103–115. doi:10.1016/s0032-5910(99)00207-7.