SlideShare a Scribd company logo
1 of 13
Download to read offline
Application of Genetic Algorithms to
describe flotation behavior in a
columnar cell
Leon, Mario1
Introduction - Hydrodynamics of Flotation
Machine
Learning
• Problems too large and too
complex.
• Unable to capture all effects for
realistic simulation.
• Constantly evolving data.
• Enough accuracy and precision
Data from the
instruments installed
in the cell column
https://benmoseley.blog/my-research/so-what-is-a-physics-informed-neural-network/
Where are:
• Hydrodynamics of
Flotation?
• Virtual sensors?
Our Approach: Physics-Informed Machine Learning
Fluid Dynamic Model - Columnar Cell Flotation
• Two-phase CFD models of a Columnar Cell
(Water + air bubbles) for the study of Gas
Holdup.
• Fluid Dynamic Models, Stokes, Sedimentation.
• Bubble density.
• Drift-Flux model for bubble diameter.
• Models based on empirical equations to validate
the models.
• Kinetic models for recovery and residence time.
• Calibration of models direct measurements (Gas
holdup).
• ...
Ikukumbuta Mwandawande, 2016
New Approach: Genetic Algorithm (GA) Application
Classical Regression
Symbolic Regression
Classical numerical regression requires the structure of the
expression to be fixed a priori
Symbolic regression allows us to obtain both the structure
of the expression and the values of the coefficients in an
automatic way
yi = β0 + β1sin(x1+x2)+ β2exp(x3) + ε
Column Cell Model
Air Flow
Slurry and wash
water Flow Rate
Froth Height
Gas Holdup
Copper Feed
grade
Collector
Dosage
Frother Dosage
Mineral and
Slurry Density
Bubble Surface
Area Flux
Recovery
Copper
Concentrate
grade
Data - Column Cells
The data was obtained from 3 different copper mines using
Columnar Cells in their flotation circuits.
For this analysis we have data from instruments installed in the
cells, the total is equivalent to a period of one year and 6
months with a frequency of 6 minutes.
The data was split into groups for training, testing and validation
in continuous data proportions of 60%, 30% and 10%.
The data was analyzed before processing in search of outliers. It
was found that 5% of the data corresponded to shutdowns,
maintenance, start-up processes, defective instruments, etc.
MEP Model - Multi-Expression Programming
• The MEP algorithm starts by creating a random
population of numerical expressions.
• Two parent individuals are selected by a selection
procedure.
• The parents recombine to obtain two offspring.
• These new offspring individuals are considered for the
mutation process.
• The best offspring will replace the worst individual in the
current population, provided it has a better fit.
• Very destructive crossover, which generates a lot of
diversity.
MEP Model Configuration
Parameter Set Value
Functions
1
Addition, subtraction, multiplication, division,
power, square root, square root, exponential,
inverse, negative, square
2
Multiplication, division, power, power, square
root, exponential, inverse, negative, square
Measurement of error 1, 2
MAE: Mean absolute error
MAD: Median absolute error
n° of sub-populations 1, 2 50
Sub-population size 1, 2 300
Tree length 1, 2 80, 40, 20
Crossover probability 1, 2 90%
Mutation probability 1, 2 1%, 5%
n° of generations 1, 2 60
n° of runs 1, 2 3000
The MEP model was processed with a
computer equipped with Xeon Processor,
4 cores, 3 GHz, 24 GB ram.
OS: Windows 10 Pro 64-bit,
Processing time one model: ~ 1 day,
n° of processed models: 10
Model
Error MAE %
Cu grade
Expression Constant
1 0.407
a0 = 0.0720
a1 = 0.3167
2 0.408 a0 = 0.0111
3 0.410
a0 = 0.3566
a2 = 0.3115
4 0.411
a0 = 0.3521
a2 = 0.0748
1
3
0
0
1
0
1
x
x
a
x
a
a
y 



3
1
0
0
1
x
x
x
a
y 


0
1
0
2
0
0
1
0
0
0
0
exp
1
a
x
a
a
x
x
x
a
a
a
x
y 























0
4
3
1
2
0
0
1
a
x
x
x
a
x
a
y 










Results
The variables are as
follows:
X0 Feed Cu grade,
X1 % solids,
X2 Wash water,
X3 Froth height,
X4 Airflow,
Y Concentrated Cu grade
Results
The Models f2, f3 y f4 have the same fit and shape as the "winning" model f1.
This expression perfectly follows the fluctuations of the measured grade as a function of the
measured variables: X0 Feed Cu grade, X1 % solids, X3 Froth height and X4 Airflow.
The models obtained with symbolic regression are numerically accurate over the range of the data
(Domain).
The best models do not contain the variable X2 Washing water, this is because this variable has a MAD /
median ratio of 5%, therefore it is close to constant.
The models do not have a "classical" structure or form of a physical model.
Analysis of the models with data outside the training range generates unexpected results, therefore,
further analysis is required to apply them to behavioral predictions.
One approach that can be applied to avoid the effect of obtaining models with "unphysical shapes" is to
use only dimensionless variables, e.g. Reynold's number, concentrate output / feed flow ratio and etc.
Conclusions
(1) The author would like to acknowledge the support of IntelliSense.io/BASF in the development of this publication.
References:
• Mihai Oltean and Crina Groşan, “Evolving digital circuits using multi expression programming”
• Crina Groşan, “Evolving mathematical expressions using Genetic Algorithms”
• Mihai Oltean and Crina Groşan, “Evolving evolutionary algorithms using multi expression programming”, 2021
• Sourabh Katoch & Sumit Singh Chauhan & Vijay Kumar, “A review on genetic algorithm: past, present, and future”,
2020
• F. Nakhaei, M. Irannajad, M. Yousefikhoshbakht, “Simultaneous optimization of flotation column performance
using genetic evolutionary algorithm”, 2016
• J.H. Holland, “Adaptation in Natural and Artificial Systems”, 1975
• Cramer, N.L., “A representation for the adaptive generation of simple sequential programs”, 1985
• Koza, J., “Genetic programming: On the programming of computers by means of natural selection”, 1992
• Mohammadzadeh S., Bolouri Bazaz , Amir H. Alavi b “An evolutionary computational approach for formulation of
compression index of fine-grained soils”, 2013
Acknowledgements / References

More Related Content

Similar to Mario Leon - IntelliSense.io - Presentacion Mineria Digital 2022.pdf

The importance of model fairness and interpretability in AI systems
The importance of model fairness and interpretability in AI systemsThe importance of model fairness and interpretability in AI systems
The importance of model fairness and interpretability in AI systemsFrancesca Lazzeri, PhD
 
Computational optimization, modelling and simulation: Recent advances and ove...
Computational optimization, modelling and simulation: Recent advances and ove...Computational optimization, modelling and simulation: Recent advances and ove...
Computational optimization, modelling and simulation: Recent advances and ove...Xin-She Yang
 
Modelling physiological uncertainty
Modelling physiological uncertaintyModelling physiological uncertainty
Modelling physiological uncertaintyNatal van Riel
 
2020.04.07 automated molecular design and the bradshaw platform webinar
2020.04.07 automated molecular design and the bradshaw platform webinar2020.04.07 automated molecular design and the bradshaw platform webinar
2020.04.07 automated molecular design and the bradshaw platform webinarPistoia Alliance
 
Applied Mathematical Modeling with Apache Solr - Joel Bernstein, Lucidworks
Applied Mathematical Modeling with Apache Solr - Joel Bernstein, LucidworksApplied Mathematical Modeling with Apache Solr - Joel Bernstein, Lucidworks
Applied Mathematical Modeling with Apache Solr - Joel Bernstein, LucidworksLucidworks
 
BIOMAG2018 - Darren Price - CamCAN
BIOMAG2018 - Darren Price - CamCANBIOMAG2018 - Darren Price - CamCAN
BIOMAG2018 - Darren Price - CamCANRobert Oostenveld
 
AIAA-SciTech-ModelSelection-2014-Mehmani
AIAA-SciTech-ModelSelection-2014-MehmaniAIAA-SciTech-ModelSelection-2014-Mehmani
AIAA-SciTech-ModelSelection-2014-MehmaniOptiModel
 
Chemoinformatic File Format.pptx
Chemoinformatic File Format.pptxChemoinformatic File Format.pptx
Chemoinformatic File Format.pptxwadhava gurumeet
 
Final Abstract Book - ICMMSND 2024_Pg. No_39.pdf
Final Abstract Book - ICMMSND 2024_Pg. No_39.pdfFinal Abstract Book - ICMMSND 2024_Pg. No_39.pdf
Final Abstract Book - ICMMSND 2024_Pg. No_39.pdfjagatheeshwari Jaga
 
Molecular modelling for in silico drug discovery
Molecular modelling for in silico drug discoveryMolecular modelling for in silico drug discovery
Molecular modelling for in silico drug discoveryLee Larcombe
 
DATI, AI E ROBOTICA @POLITO
DATI, AI E ROBOTICA @POLITODATI, AI E ROBOTICA @POLITO
DATI, AI E ROBOTICA @POLITOMarcoMellia
 
Online learning in estimation of distribution algorithms for dynamic environm...
Online learning in estimation of distribution algorithms for dynamic environm...Online learning in estimation of distribution algorithms for dynamic environm...
Online learning in estimation of distribution algorithms for dynamic environm...André Gonçalves
 
12918 2015 article_144 (1)
12918 2015 article_144 (1)12918 2015 article_144 (1)
12918 2015 article_144 (1)Anandsingh06
 
Advanced Optimization Techniques
Advanced Optimization TechniquesAdvanced Optimization Techniques
Advanced Optimization TechniquesValerie Felton
 
Diagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography DataDiagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography DataTeruKamogashira
 
Computational Fluid Dynamics (CFD)
Computational Fluid Dynamics (CFD)Computational Fluid Dynamics (CFD)
Computational Fluid Dynamics (CFD)Khusro Kamaluddin
 
Parallelization of Yeast Diode and Implementation of a Concentration Gradient
Parallelization of Yeast Diode and Implementation of a Concentration GradientParallelization of Yeast Diode and Implementation of a Concentration Gradient
Parallelization of Yeast Diode and Implementation of a Concentration GradientDouglas Cohen
 
Machine Learning and AI: Core Methods and Applications
Machine Learning and AI: Core Methods and ApplicationsMachine Learning and AI: Core Methods and Applications
Machine Learning and AI: Core Methods and ApplicationsQuantUniversity
 

Similar to Mario Leon - IntelliSense.io - Presentacion Mineria Digital 2022.pdf (20)

The importance of model fairness and interpretability in AI systems
The importance of model fairness and interpretability in AI systemsThe importance of model fairness and interpretability in AI systems
The importance of model fairness and interpretability in AI systems
 
Computational optimization, modelling and simulation: Recent advances and ove...
Computational optimization, modelling and simulation: Recent advances and ove...Computational optimization, modelling and simulation: Recent advances and ove...
Computational optimization, modelling and simulation: Recent advances and ove...
 
Modelling physiological uncertainty
Modelling physiological uncertaintyModelling physiological uncertainty
Modelling physiological uncertainty
 
2020.04.07 automated molecular design and the bradshaw platform webinar
2020.04.07 automated molecular design and the bradshaw platform webinar2020.04.07 automated molecular design and the bradshaw platform webinar
2020.04.07 automated molecular design and the bradshaw platform webinar
 
Applied Mathematical Modeling with Apache Solr - Joel Bernstein, Lucidworks
Applied Mathematical Modeling with Apache Solr - Joel Bernstein, LucidworksApplied Mathematical Modeling with Apache Solr - Joel Bernstein, Lucidworks
Applied Mathematical Modeling with Apache Solr - Joel Bernstein, Lucidworks
 
BIOMAG2018 - Darren Price - CamCAN
BIOMAG2018 - Darren Price - CamCANBIOMAG2018 - Darren Price - CamCAN
BIOMAG2018 - Darren Price - CamCAN
 
AIAA-SciTech-ModelSelection-2014-Mehmani
AIAA-SciTech-ModelSelection-2014-MehmaniAIAA-SciTech-ModelSelection-2014-Mehmani
AIAA-SciTech-ModelSelection-2014-Mehmani
 
Chemoinformatic File Format.pptx
Chemoinformatic File Format.pptxChemoinformatic File Format.pptx
Chemoinformatic File Format.pptx
 
Daniel Lee STAN
Daniel Lee STANDaniel Lee STAN
Daniel Lee STAN
 
Final Abstract Book - ICMMSND 2024_Pg. No_39.pdf
Final Abstract Book - ICMMSND 2024_Pg. No_39.pdfFinal Abstract Book - ICMMSND 2024_Pg. No_39.pdf
Final Abstract Book - ICMMSND 2024_Pg. No_39.pdf
 
call for papers, research paper publishing, where to publish research paper, ...
call for papers, research paper publishing, where to publish research paper, ...call for papers, research paper publishing, where to publish research paper, ...
call for papers, research paper publishing, where to publish research paper, ...
 
Molecular modelling for in silico drug discovery
Molecular modelling for in silico drug discoveryMolecular modelling for in silico drug discovery
Molecular modelling for in silico drug discovery
 
DATI, AI E ROBOTICA @POLITO
DATI, AI E ROBOTICA @POLITODATI, AI E ROBOTICA @POLITO
DATI, AI E ROBOTICA @POLITO
 
Online learning in estimation of distribution algorithms for dynamic environm...
Online learning in estimation of distribution algorithms for dynamic environm...Online learning in estimation of distribution algorithms for dynamic environm...
Online learning in estimation of distribution algorithms for dynamic environm...
 
12918 2015 article_144 (1)
12918 2015 article_144 (1)12918 2015 article_144 (1)
12918 2015 article_144 (1)
 
Advanced Optimization Techniques
Advanced Optimization TechniquesAdvanced Optimization Techniques
Advanced Optimization Techniques
 
Diagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography DataDiagnosis Support by Machine Learning Using Posturography Data
Diagnosis Support by Machine Learning Using Posturography Data
 
Computational Fluid Dynamics (CFD)
Computational Fluid Dynamics (CFD)Computational Fluid Dynamics (CFD)
Computational Fluid Dynamics (CFD)
 
Parallelization of Yeast Diode and Implementation of a Concentration Gradient
Parallelization of Yeast Diode and Implementation of a Concentration GradientParallelization of Yeast Diode and Implementation of a Concentration Gradient
Parallelization of Yeast Diode and Implementation of a Concentration Gradient
 
Machine Learning and AI: Core Methods and Applications
Machine Learning and AI: Core Methods and ApplicationsMachine Learning and AI: Core Methods and Applications
Machine Learning and AI: Core Methods and Applications
 

Recently uploaded

Passive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.pptPassive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.pptamrabdallah9
 
Seizure stage detection of epileptic seizure using convolutional neural networks
Seizure stage detection of epileptic seizure using convolutional neural networksSeizure stage detection of epileptic seizure using convolutional neural networks
Seizure stage detection of epileptic seizure using convolutional neural networksIJECEIAES
 
8th International Conference on Soft Computing, Mathematics and Control (SMC ...
8th International Conference on Soft Computing, Mathematics and Control (SMC ...8th International Conference on Soft Computing, Mathematics and Control (SMC ...
8th International Conference on Soft Computing, Mathematics and Control (SMC ...josephjonse
 
ALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdfALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdfMadan Karki
 
Filters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsFilters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsMathias Magdowski
 
The Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptxThe Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptxMANASINANDKISHORDEOR
 
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdfInvolute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdfJNTUA
 
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdflitvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdfAlexander Litvinenko
 
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas SachpazisSeismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas SachpazisDr.Costas Sachpazis
 
Online crime reporting system project.pdf
Online crime reporting system project.pdfOnline crime reporting system project.pdf
Online crime reporting system project.pdfKamal Acharya
 
Worksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptxWorksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptxMustafa Ahmed
 
Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...
Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...
Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...drjose256
 
Autodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptxAutodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptxMustafa Ahmed
 
Instruct Nirmaana 24-Smart and Lean Construction Through Technology.pdf
Instruct Nirmaana 24-Smart and Lean Construction Through Technology.pdfInstruct Nirmaana 24-Smart and Lean Construction Through Technology.pdf
Instruct Nirmaana 24-Smart and Lean Construction Through Technology.pdfEr.Sonali Nasikkar
 
Operating System chapter 9 (Virtual Memory)
Operating System chapter 9 (Virtual Memory)Operating System chapter 9 (Virtual Memory)
Operating System chapter 9 (Virtual Memory)NareenAsad
 
Low Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s HandbookLow Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s HandbookPeterJack13
 
Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2T.D. Shashikala
 
NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...
NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...
NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...Amil baba
 
Diploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdfDiploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdfJNTUA
 
Raashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashidFaiyazSheikh
 

Recently uploaded (20)

Passive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.pptPassive Air Cooling System and Solar Water Heater.ppt
Passive Air Cooling System and Solar Water Heater.ppt
 
Seizure stage detection of epileptic seizure using convolutional neural networks
Seizure stage detection of epileptic seizure using convolutional neural networksSeizure stage detection of epileptic seizure using convolutional neural networks
Seizure stage detection of epileptic seizure using convolutional neural networks
 
8th International Conference on Soft Computing, Mathematics and Control (SMC ...
8th International Conference on Soft Computing, Mathematics and Control (SMC ...8th International Conference on Soft Computing, Mathematics and Control (SMC ...
8th International Conference on Soft Computing, Mathematics and Control (SMC ...
 
ALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdfALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdf
 
Filters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsFilters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility Applications
 
The Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptxThe Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptx
 
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdfInvolute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
Involute of a circle,Square, pentagon,HexagonInvolute_Engineering Drawing.pdf
 
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdflitvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
 
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas SachpazisSeismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
 
Online crime reporting system project.pdf
Online crime reporting system project.pdfOnline crime reporting system project.pdf
Online crime reporting system project.pdf
 
Worksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptxWorksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptx
 
Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...
Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...
Tembisa Central Terminating Pills +27838792658 PHOMOLONG Top Abortion Pills F...
 
Autodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptxAutodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptx
 
Instruct Nirmaana 24-Smart and Lean Construction Through Technology.pdf
Instruct Nirmaana 24-Smart and Lean Construction Through Technology.pdfInstruct Nirmaana 24-Smart and Lean Construction Through Technology.pdf
Instruct Nirmaana 24-Smart and Lean Construction Through Technology.pdf
 
Operating System chapter 9 (Virtual Memory)
Operating System chapter 9 (Virtual Memory)Operating System chapter 9 (Virtual Memory)
Operating System chapter 9 (Virtual Memory)
 
Low Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s HandbookLow Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s Handbook
 
Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2
 
NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...
NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...
NO1 Best Powerful Vashikaran Specialist Baba Vashikaran Specialist For Love V...
 
Diploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdfDiploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdf
 
Raashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashid final report on Embedded Systems
Raashid final report on Embedded Systems
 

Mario Leon - IntelliSense.io - Presentacion Mineria Digital 2022.pdf

  • 1. Application of Genetic Algorithms to describe flotation behavior in a columnar cell Leon, Mario1
  • 2. Introduction - Hydrodynamics of Flotation Machine Learning • Problems too large and too complex. • Unable to capture all effects for realistic simulation. • Constantly evolving data. • Enough accuracy and precision Data from the instruments installed in the cell column https://benmoseley.blog/my-research/so-what-is-a-physics-informed-neural-network/ Where are: • Hydrodynamics of Flotation? • Virtual sensors?
  • 4. Fluid Dynamic Model - Columnar Cell Flotation • Two-phase CFD models of a Columnar Cell (Water + air bubbles) for the study of Gas Holdup. • Fluid Dynamic Models, Stokes, Sedimentation. • Bubble density. • Drift-Flux model for bubble diameter. • Models based on empirical equations to validate the models. • Kinetic models for recovery and residence time. • Calibration of models direct measurements (Gas holdup). • ... Ikukumbuta Mwandawande, 2016
  • 5. New Approach: Genetic Algorithm (GA) Application Classical Regression Symbolic Regression Classical numerical regression requires the structure of the expression to be fixed a priori Symbolic regression allows us to obtain both the structure of the expression and the values of the coefficients in an automatic way yi = β0 + β1sin(x1+x2)+ β2exp(x3) + ε
  • 6. Column Cell Model Air Flow Slurry and wash water Flow Rate Froth Height Gas Holdup Copper Feed grade Collector Dosage Frother Dosage Mineral and Slurry Density Bubble Surface Area Flux Recovery Copper Concentrate grade
  • 7. Data - Column Cells The data was obtained from 3 different copper mines using Columnar Cells in their flotation circuits. For this analysis we have data from instruments installed in the cells, the total is equivalent to a period of one year and 6 months with a frequency of 6 minutes. The data was split into groups for training, testing and validation in continuous data proportions of 60%, 30% and 10%. The data was analyzed before processing in search of outliers. It was found that 5% of the data corresponded to shutdowns, maintenance, start-up processes, defective instruments, etc.
  • 8. MEP Model - Multi-Expression Programming • The MEP algorithm starts by creating a random population of numerical expressions. • Two parent individuals are selected by a selection procedure. • The parents recombine to obtain two offspring. • These new offspring individuals are considered for the mutation process. • The best offspring will replace the worst individual in the current population, provided it has a better fit. • Very destructive crossover, which generates a lot of diversity.
  • 9. MEP Model Configuration Parameter Set Value Functions 1 Addition, subtraction, multiplication, division, power, square root, square root, exponential, inverse, negative, square 2 Multiplication, division, power, power, square root, exponential, inverse, negative, square Measurement of error 1, 2 MAE: Mean absolute error MAD: Median absolute error n° of sub-populations 1, 2 50 Sub-population size 1, 2 300 Tree length 1, 2 80, 40, 20 Crossover probability 1, 2 90% Mutation probability 1, 2 1%, 5% n° of generations 1, 2 60 n° of runs 1, 2 3000 The MEP model was processed with a computer equipped with Xeon Processor, 4 cores, 3 GHz, 24 GB ram. OS: Windows 10 Pro 64-bit, Processing time one model: ~ 1 day, n° of processed models: 10
  • 10. Model Error MAE % Cu grade Expression Constant 1 0.407 a0 = 0.0720 a1 = 0.3167 2 0.408 a0 = 0.0111 3 0.410 a0 = 0.3566 a2 = 0.3115 4 0.411 a0 = 0.3521 a2 = 0.0748 1 3 0 0 1 0 1 x x a x a a y     3 1 0 0 1 x x x a y    0 1 0 2 0 0 1 0 0 0 0 exp 1 a x a a x x x a a a x y                         0 4 3 1 2 0 0 1 a x x x a x a y            Results The variables are as follows: X0 Feed Cu grade, X1 % solids, X2 Wash water, X3 Froth height, X4 Airflow, Y Concentrated Cu grade
  • 11. Results The Models f2, f3 y f4 have the same fit and shape as the "winning" model f1. This expression perfectly follows the fluctuations of the measured grade as a function of the measured variables: X0 Feed Cu grade, X1 % solids, X3 Froth height and X4 Airflow.
  • 12. The models obtained with symbolic regression are numerically accurate over the range of the data (Domain). The best models do not contain the variable X2 Washing water, this is because this variable has a MAD / median ratio of 5%, therefore it is close to constant. The models do not have a "classical" structure or form of a physical model. Analysis of the models with data outside the training range generates unexpected results, therefore, further analysis is required to apply them to behavioral predictions. One approach that can be applied to avoid the effect of obtaining models with "unphysical shapes" is to use only dimensionless variables, e.g. Reynold's number, concentrate output / feed flow ratio and etc. Conclusions
  • 13. (1) The author would like to acknowledge the support of IntelliSense.io/BASF in the development of this publication. References: • Mihai Oltean and Crina Groşan, “Evolving digital circuits using multi expression programming” • Crina Groşan, “Evolving mathematical expressions using Genetic Algorithms” • Mihai Oltean and Crina Groşan, “Evolving evolutionary algorithms using multi expression programming”, 2021 • Sourabh Katoch & Sumit Singh Chauhan & Vijay Kumar, “A review on genetic algorithm: past, present, and future”, 2020 • F. Nakhaei, M. Irannajad, M. Yousefikhoshbakht, “Simultaneous optimization of flotation column performance using genetic evolutionary algorithm”, 2016 • J.H. Holland, “Adaptation in Natural and Artificial Systems”, 1975 • Cramer, N.L., “A representation for the adaptive generation of simple sequential programs”, 1985 • Koza, J., “Genetic programming: On the programming of computers by means of natural selection”, 1992 • Mohammadzadeh S., Bolouri Bazaz , Amir H. Alavi b “An evolutionary computational approach for formulation of compression index of fine-grained soils”, 2013 Acknowledgements / References