SlideShare a Scribd company logo
1 of 36
Download to read offline
Christiaan Erdbrink 
chriserdbrink@gmail.com 
Data Science Symposium 
31.10.2014 
System identification 
using evolutionary computing
My background 
MSc Delft University of Technology, 
Civil Engineering, fluid mechanics 
Deltares, flow around hydraulic structures 
PhD University of Amsterdam, 
Computational Science 
Saint Petersburg National Research University of Information Technologies, Mechanics and Optics (ITMO University)
Christiaan Erdbrink 
Problem description 
Solution strategies 
- traditional 
- CFD 
- data-driven 
Evolutionary computing 
Conclusions & Outlook on future work 
Questions/Discussion 
Outline
Christiaan Erdbrink 
nl.wikipedia.org 
infopuntveiligheid.nl 
coflexbouweninfra.noordhoff.nl 
Problem description
Christiaan Erdbrink 
Haringvliet barrier 
Problem description
Christiaan Erdbrink 
Physics of flow-induced vibrations 
(in a nutshell) 
Problem description 
Excitation mechanisms 
turbulence 
stable vortex shedding 
flow instabilities 
self-excitation 
unstable fluid resonance
Christiaan Erdbrink 
Physics of flow-induced vibrations 
(in a nutshell) 
Problem description 
observed response 
excitation mechanisms 
assessment 
measures 
gate design 
real-life conditions
Christiaan Erdbrink 
Problem description 
Solution strategies 
- traditional 
- CFD 
- data-driven 
Evolutionary computing 
Conclusions & Outlook on future work 
Questions/Discussion
Christiaan Erdbrink 
Haringvliet barrier 
Traditional solutions
Christiaan Erdbrink 
gate 
Traditional solutions 
Erdbrink, Krzhizhanovskaya, Sloot (2014): 
“Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids & Structures.
Christiaan Erdbrink 
Traditional solutions
Christiaan Erdbrink 
A/D = f( ζ , mr , Vr , Fr, St, I ) 
f a Δh Cs 
Traditional solutions 
Vr (-) 
Fz 
(-) 
Erdbrink, Krzhizhanovskaya, Sloot (2014): 
“Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids & Structures.
Christiaan Erdbrink 
xi (t)  xi (t)  xi 
a 
Δh 
f 
Traditional solutions
Christiaan Erdbrink 
Problem description 
Solution strategies 
- traditional 
- CFD 
- data-driven 
Evolutionary computing 
Conclusions & Outlook on future work 
Questions/Discussion
Christiaan Erdbrink 
gate with leakage 
numerical simulations: CFD 
Erdbrink, Krzhizhanovskaya, Sloot (2014): 
“Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids and Structures.
Christiaan Erdbrink 
at Vr ≈ 10: 
numerical simulations: CFD 
Erdbrink, Krzhizhanovskaya, Sloot (2014): 
“Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids and Structures.
Christiaan Erdbrink 
Problem description 
Solution strategies 
- traditional 
- CFD 
- data-driven 
Evolutionary computing 
Conclusions & Outlook on future work 
Questions/Discussion
Christiaan Erdbrink 
h1(t) 
h2(t) 
safe or unsafe 
data-driven solution 
xi (t)  xi (t)  xi (t) 
a(t) 
f(t) 
xcrit
Christiaan Erdbrink 
Use classification to avoid critical regions 
data-driven solution 
Vr (-) 
a (m) 
Erdbrink, Krzhizhanovskaya, Sloot (2012): 
“Controlling flow-induced vibrations of flood barrier gates with 
data-driven and finite-element modelling”, FLOODrisk2012
Christiaan Erdbrink 
Designing a control system 
data-driven solution
Christiaan Erdbrink 
Problem description 
Solution strategies 
- traditional 
- CFD 
- data-driven 
Evolutionary computing 
Outlook & Conclusions 
Questions/Discussion
Christiaan Erdbrink 
approaches 
evolutionary computing 
traditional 
field measurements 
physical modelling 
FIV problems 
numerical simulations 
data-driven 
for control: 
for system id: 
classification 
evolutionary computing 
signal analysis 
FEM 
CFD, 
CFSI 
classic 
machine learning 
differential evolution 
genetic programming
Christiaan Erdbrink 
evolutionary computing 
Hornby et al. (2006): “Automated antenna design with evolutionary algorithms” 
en.wikipedia.org/wiki/Evolved_antenna
Christiaan Erdbrink 
Evolutionary algorithms 
Eiben & Smith (2011): 
“Introduction to evolutionary computing” 
evolutionary computing
Christiaan Erdbrink 
Evolutionary Computing 
Meta-heuristics 
evolutionary computing
Christiaan Erdbrink 
In general, EAs work well: - for multimodal problems - for multi-objective optimization - in hard design problems where a proposed configuration can be tested unambiguously - when small improvements are appreciated - when speed is not essential - when standard methods fail 
evolutionary computing
Christiaan Erdbrink 
Reverse engineering dynamical systems 
For example, 
evolutionary computing 
Erdbrink, Krzhizhanovskaya: 
“Identifying Self‐Excited Vibrations with Evolutionary Computing”, Procedia Computer Science, Vol.29, pp.637‐647. 
Sensitivity analyses 
population size 
termination 
model parameters 
evaluation tolerance 
solver type
Christiaan Erdbrink 
Fitness progression 
- model parameters 
updated once in 20 gens 
updated each gen 
evolutionary computing
Christiaan Erdbrink 
0 2 4 6 8 10 
x 10 
4 
2 
4 
6 
8 
10 
12 
generation 
fitness 
best 
Fitness progression 
- termination criterion 
0 500 1000 1500 
2 
4 
6 
8 
10 
12 
generation 
fitness 
best 
evolutionary computing
Christiaan Erdbrink 
Genetic Programming – applied to Symbolic Regression 
(x-C)*log(2x) 
y = 
(x-2.7139)*log(2x) 
*-LxC+xx 
evolutionary computing 
y = f(x,y) 
, etc.
Christiaan Erdbrink 
Example 
y = 0.468e|x|sin(3.795x) 
evolutionary computing
Christiaan Erdbrink 
evolutionary computing 
Mining physical systems (the “robot scientist”) 
M Schmidt, and H Lipson Science 2009;324:81-85
Christiaan Erdbrink 
Application example 
evolutionary computing 
C.D. Erdbrink (2014): 
“Modelling flow-induced vibrations of gates in hydraulic structures”, PhD thesis Univ. of Amsterdam
Christiaan Erdbrink 
Problem description 
Solution strategies 
- traditional 
- CFD 
- data-driven 
Evolutionary computing 
Outlook & Conclusions 
Questions/Discussion
Christiaan Erdbrink 
Conclusions 
Data-driven methods should not be seen as competitors of traditional forms of modelling, but as valuable complementary tools. 
Monitoring of gate behaviour combined with classification of dynamic response states can be used to avoid critical vibration ranges. 
Evolutionary Computing… 
…is a versatile approach for all kinds of optimization problems. 
…has evolved from a hobby for computer scientists to an important area of research, with innumerous successful applications. 
…can be applied to output-only identification of (complex) dynamical systems. 
…is capable of automatically deriving meaningful elementary equations and from data.
Christiaan Erdbrink 
chriserdbrink@gmail.com 
Thank you for your attention!

More Related Content

Viewers also liked

Creators I blog because I have something important to share --StevenNoble
Creators I blog because I have something important to share --StevenNobleCreators I blog because I have something important to share --StevenNoble
Creators I blog because I have something important to share --StevenNobleadtech
 
The year of mobility
The year of mobility The year of mobility
The year of mobility adtech
 
Trigger Based Email How to add value to your marketing mix
Trigger Based Email How to add value to your marketing mix Trigger Based Email How to add value to your marketing mix
Trigger Based Email How to add value to your marketing mix adtech
 
The Transf orm ati on
The Transf orm ati on The Transf orm ati on
The Transf orm ati on adtech
 
The new web: mobile New technologies, new applications (new ways of connectin...
The new web: mobile New technologies, new applications (new ways of connectin...The new web: mobile New technologies, new applications (new ways of connectin...
The new web: mobile New technologies, new applications (new ways of connectin...adtech
 
# Handing your brand to consumers Addressing the risks and realising the bene...
# Handing your brand to consumers Addressing the risks and realising the bene...# Handing your brand to consumers Addressing the risks and realising the bene...
# Handing your brand to consumers Addressing the risks and realising the bene...adtech
 
Habilidades sociales 3
Habilidades sociales 3Habilidades sociales 3
Habilidades sociales 3Emagister
 
UC Berkeley Data Science Webinar
UC Berkeley Data Science WebinarUC Berkeley Data Science Webinar
UC Berkeley Data Science WebinarAlpine Data
 
Best Practices for a Data-driven Approach to Test Utilization
Best Practices for a Data-driven Approach to Test UtilizationBest Practices for a Data-driven Approach to Test Utilization
Best Practices for a Data-driven Approach to Test UtilizationViewics
 
Acquisition enews registration process
Acquisition enews registration process Acquisition enews registration process
Acquisition enews registration process adtech
 
Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...
Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...
Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...Ontico
 

Viewers also liked (13)

7142
71427142
7142
 
Creators I blog because I have something important to share --StevenNoble
Creators I blog because I have something important to share --StevenNobleCreators I blog because I have something important to share --StevenNoble
Creators I blog because I have something important to share --StevenNoble
 
The year of mobility
The year of mobility The year of mobility
The year of mobility
 
Trigger Based Email How to add value to your marketing mix
Trigger Based Email How to add value to your marketing mix Trigger Based Email How to add value to your marketing mix
Trigger Based Email How to add value to your marketing mix
 
The Transf orm ati on
The Transf orm ati on The Transf orm ati on
The Transf orm ati on
 
The new web: mobile New technologies, new applications (new ways of connectin...
The new web: mobile New technologies, new applications (new ways of connectin...The new web: mobile New technologies, new applications (new ways of connectin...
The new web: mobile New technologies, new applications (new ways of connectin...
 
# Handing your brand to consumers Addressing the risks and realising the bene...
# Handing your brand to consumers Addressing the risks and realising the bene...# Handing your brand to consumers Addressing the risks and realising the bene...
# Handing your brand to consumers Addressing the risks and realising the bene...
 
Habilidades sociales 3
Habilidades sociales 3Habilidades sociales 3
Habilidades sociales 3
 
UC Berkeley Data Science Webinar
UC Berkeley Data Science WebinarUC Berkeley Data Science Webinar
UC Berkeley Data Science Webinar
 
Best Practices for a Data-driven Approach to Test Utilization
Best Practices for a Data-driven Approach to Test UtilizationBest Practices for a Data-driven Approach to Test Utilization
Best Practices for a Data-driven Approach to Test Utilization
 
Viruses
VirusesViruses
Viruses
 
Acquisition enews registration process
Acquisition enews registration process Acquisition enews registration process
Acquisition enews registration process
 
Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...
Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...
Аналитика мобильного проекта — проверяй и доверяй / Александр Лукин (Yandex A...
 

Similar to DSD-INT 2014 - Data Science symposium - Application 2 - System identification with evolutionary computing, Dr. Christiaan Erdbrink, University of Amsterdam

Mit2 092 f09_lec01
Mit2 092 f09_lec01Mit2 092 f09_lec01
Mit2 092 f09_lec01Rahman Hakim
 
Computational fluid dynamics
Computational fluid dynamicsComputational fluid dynamics
Computational fluid dynamicsRavi Choudhary
 
Multidisciplinary analysis and optimization under uncertainty
Multidisciplinary analysis and optimization under uncertaintyMultidisciplinary analysis and optimization under uncertainty
Multidisciplinary analysis and optimization under uncertaintyChen Liang
 
The study on mining temporal patterns and related applications in dynamic soc...
The study on mining temporal patterns and related applications in dynamic soc...The study on mining temporal patterns and related applications in dynamic soc...
The study on mining temporal patterns and related applications in dynamic soc...Thanh Hieu
 
From Classroom to Collaboration: Crossing Computational and Classic Chemistry
From Classroom to Collaboration: Crossing Computational and Classic ChemistryFrom Classroom to Collaboration: Crossing Computational and Classic Chemistry
From Classroom to Collaboration: Crossing Computational and Classic Chemistrykarl.barnes
 
Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...
Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...
Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...Chris Rackauckas
 
CE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element Analysis
CE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element AnalysisCE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element Analysis
CE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element AnalysisFawad Najam
 
Streaming Model Transformations by Complex Event Processing
Streaming Model Transformations by Complex Event ProcessingStreaming Model Transformations by Complex Event Processing
Streaming Model Transformations by Complex Event ProcessingIstván Dávid
 
Evolution of Graph Algorithms – Benefits and Challenges
Evolution of Graph Algorithms – Benefits and ChallengesEvolution of Graph Algorithms – Benefits and Challenges
Evolution of Graph Algorithms – Benefits and ChallengesEbru Cucen Çüçen
 
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...PyData
 
Automated seismic-to-well ties?
Automated seismic-to-well ties?Automated seismic-to-well ties?
Automated seismic-to-well ties?UT Technology
 
Mpp Rsv 2008 Public
Mpp Rsv 2008 PublicMpp Rsv 2008 Public
Mpp Rsv 2008 Publiclab13unisa
 
2.ME VLSI 233-251.doc
2.ME VLSI 233-251.doc2.ME VLSI 233-251.doc
2.ME VLSI 233-251.docSRI NISHITH
 
"An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ..."An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ...butest
 
Metaheuristics and Optimiztion in Civil Engineering
Metaheuristics and Optimiztion in Civil EngineeringMetaheuristics and Optimiztion in Civil Engineering
Metaheuristics and Optimiztion in Civil EngineeringXin-She Yang
 
M3research 2023 (1).pdf
M3research 2023 (1).pdfM3research 2023 (1).pdf
M3research 2023 (1).pdfKldAli2
 
Aerodynamic design of Aircraft”
Aerodynamic design of Aircraft”Aerodynamic design of Aircraft”
Aerodynamic design of Aircraft”Masahiro Kanazaki
 
Multiphysics Group at HSR
Multiphysics Group at HSRMultiphysics Group at HSR
Multiphysics Group at HSRmictc
 

Similar to DSD-INT 2014 - Data Science symposium - Application 2 - System identification with evolutionary computing, Dr. Christiaan Erdbrink, University of Amsterdam (20)

Mit2 092 f09_lec01
Mit2 092 f09_lec01Mit2 092 f09_lec01
Mit2 092 f09_lec01
 
Computational fluid dynamics
Computational fluid dynamicsComputational fluid dynamics
Computational fluid dynamics
 
Multidisciplinary analysis and optimization under uncertainty
Multidisciplinary analysis and optimization under uncertaintyMultidisciplinary analysis and optimization under uncertainty
Multidisciplinary analysis and optimization under uncertainty
 
The study on mining temporal patterns and related applications in dynamic soc...
The study on mining temporal patterns and related applications in dynamic soc...The study on mining temporal patterns and related applications in dynamic soc...
The study on mining temporal patterns and related applications in dynamic soc...
 
From Classroom to Collaboration: Crossing Computational and Classic Chemistry
From Classroom to Collaboration: Crossing Computational and Classic ChemistryFrom Classroom to Collaboration: Crossing Computational and Classic Chemistry
From Classroom to Collaboration: Crossing Computational and Classic Chemistry
 
Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...
Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...
Generalizing Scientific Machine Learning and Differentiable Simulation Beyond...
 
CE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element Analysis
CE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element AnalysisCE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element Analysis
CE 72.32 (January 2016 Semester) Lecture 6 - Overview of Finite Element Analysis
 
Streaming Model Transformations by Complex Event Processing
Streaming Model Transformations by Complex Event ProcessingStreaming Model Transformations by Complex Event Processing
Streaming Model Transformations by Complex Event Processing
 
Evolution of Graph Algorithms – Benefits and Challenges
Evolution of Graph Algorithms – Benefits and ChallengesEvolution of Graph Algorithms – Benefits and Challenges
Evolution of Graph Algorithms – Benefits and Challenges
 
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
 
Automated seismic-to-well ties?
Automated seismic-to-well ties?Automated seismic-to-well ties?
Automated seismic-to-well ties?
 
Mpp Rsv 2008 Public
Mpp Rsv 2008 PublicMpp Rsv 2008 Public
Mpp Rsv 2008 Public
 
2.ME VLSI 233-251.doc
2.ME VLSI 233-251.doc2.ME VLSI 233-251.doc
2.ME VLSI 233-251.doc
 
"An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ..."An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ...
 
Metaheuristics and Optimiztion in Civil Engineering
Metaheuristics and Optimiztion in Civil EngineeringMetaheuristics and Optimiztion in Civil Engineering
Metaheuristics and Optimiztion in Civil Engineering
 
M3research 2023 (1).pdf
M3research 2023 (1).pdfM3research 2023 (1).pdf
M3research 2023 (1).pdf
 
Isav2012 draft1final (1)
Isav2012 draft1final (1)Isav2012 draft1final (1)
Isav2012 draft1final (1)
 
Aerodynamic design of Aircraft”
Aerodynamic design of Aircraft”Aerodynamic design of Aircraft”
Aerodynamic design of Aircraft”
 
Multiphysics Group at HSR
Multiphysics Group at HSRMultiphysics Group at HSR
Multiphysics Group at HSR
 
Phd Defense 2007
Phd Defense 2007Phd Defense 2007
Phd Defense 2007
 

Recently uploaded

Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43bNightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43bSérgio Sacani
 
Physiochemical properties of nanomaterials and its nanotoxicity.pptx
Physiochemical properties of nanomaterials and its nanotoxicity.pptxPhysiochemical properties of nanomaterials and its nanotoxicity.pptx
Physiochemical properties of nanomaterials and its nanotoxicity.pptxAArockiyaNisha
 
Biological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdfBiological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdfmuntazimhurra
 
Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...
Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...
Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...anilsa9823
 
TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...
TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...
TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...ssifa0344
 
GBSN - Microbiology (Unit 1)
GBSN - Microbiology (Unit 1)GBSN - Microbiology (Unit 1)
GBSN - Microbiology (Unit 1)Areesha Ahmad
 
Pests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdfPests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdfPirithiRaju
 
Chromatin Structure | EUCHROMATIN | HETEROCHROMATIN
Chromatin Structure | EUCHROMATIN | HETEROCHROMATINChromatin Structure | EUCHROMATIN | HETEROCHROMATIN
Chromatin Structure | EUCHROMATIN | HETEROCHROMATINsankalpkumarsahoo174
 
Disentangling the origin of chemical differences using GHOST
Disentangling the origin of chemical differences using GHOSTDisentangling the origin of chemical differences using GHOST
Disentangling the origin of chemical differences using GHOSTSérgio Sacani
 
Recombination DNA Technology (Nucleic Acid Hybridization )
Recombination DNA Technology (Nucleic Acid Hybridization )Recombination DNA Technology (Nucleic Acid Hybridization )
Recombination DNA Technology (Nucleic Acid Hybridization )aarthirajkumar25
 
Formation of low mass protostars and their circumstellar disks
Formation of low mass protostars and their circumstellar disksFormation of low mass protostars and their circumstellar disks
Formation of low mass protostars and their circumstellar disksSérgio Sacani
 
fundamental of entomology all in one topics of entomology
fundamental of entomology all in one topics of entomologyfundamental of entomology all in one topics of entomology
fundamental of entomology all in one topics of entomologyDrAnita Sharma
 
Stunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCR
Stunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCRStunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCR
Stunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCRDelhi Call girls
 
Hubble Asteroid Hunter III. Physical properties of newly found asteroids
Hubble Asteroid Hunter III. Physical properties of newly found asteroidsHubble Asteroid Hunter III. Physical properties of newly found asteroids
Hubble Asteroid Hunter III. Physical properties of newly found asteroidsSérgio Sacani
 
Pests of cotton_Borer_Pests_Binomics_Dr.UPR.pdf
Pests of cotton_Borer_Pests_Binomics_Dr.UPR.pdfPests of cotton_Borer_Pests_Binomics_Dr.UPR.pdf
Pests of cotton_Borer_Pests_Binomics_Dr.UPR.pdfPirithiRaju
 
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 60009654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000Sapana Sha
 
Presentation Vikram Lander by Vedansh Gupta.pptx
Presentation Vikram Lander by Vedansh Gupta.pptxPresentation Vikram Lander by Vedansh Gupta.pptx
Presentation Vikram Lander by Vedansh Gupta.pptxgindu3009
 
DIFFERENCE IN BACK CROSS AND TEST CROSS
DIFFERENCE IN  BACK CROSS AND TEST CROSSDIFFERENCE IN  BACK CROSS AND TEST CROSS
DIFFERENCE IN BACK CROSS AND TEST CROSSLeenakshiTyagi
 
GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)Areesha Ahmad
 

Recently uploaded (20)

Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43bNightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
 
Physiochemical properties of nanomaterials and its nanotoxicity.pptx
Physiochemical properties of nanomaterials and its nanotoxicity.pptxPhysiochemical properties of nanomaterials and its nanotoxicity.pptx
Physiochemical properties of nanomaterials and its nanotoxicity.pptx
 
Biological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdfBiological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdf
 
Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...
Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...
Lucknow 💋 Russian Call Girls Lucknow Finest Escorts Service 8923113531 Availa...
 
TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...
TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...
TEST BANK For Radiologic Science for Technologists, 12th Edition by Stewart C...
 
GBSN - Microbiology (Unit 1)
GBSN - Microbiology (Unit 1)GBSN - Microbiology (Unit 1)
GBSN - Microbiology (Unit 1)
 
Pests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdfPests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdf
 
The Philosophy of Science
The Philosophy of ScienceThe Philosophy of Science
The Philosophy of Science
 
Chromatin Structure | EUCHROMATIN | HETEROCHROMATIN
Chromatin Structure | EUCHROMATIN | HETEROCHROMATINChromatin Structure | EUCHROMATIN | HETEROCHROMATIN
Chromatin Structure | EUCHROMATIN | HETEROCHROMATIN
 
Disentangling the origin of chemical differences using GHOST
Disentangling the origin of chemical differences using GHOSTDisentangling the origin of chemical differences using GHOST
Disentangling the origin of chemical differences using GHOST
 
Recombination DNA Technology (Nucleic Acid Hybridization )
Recombination DNA Technology (Nucleic Acid Hybridization )Recombination DNA Technology (Nucleic Acid Hybridization )
Recombination DNA Technology (Nucleic Acid Hybridization )
 
Formation of low mass protostars and their circumstellar disks
Formation of low mass protostars and their circumstellar disksFormation of low mass protostars and their circumstellar disks
Formation of low mass protostars and their circumstellar disks
 
fundamental of entomology all in one topics of entomology
fundamental of entomology all in one topics of entomologyfundamental of entomology all in one topics of entomology
fundamental of entomology all in one topics of entomology
 
Stunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCR
Stunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCRStunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCR
Stunning ➥8448380779▻ Call Girls In Panchshil Enclave Delhi NCR
 
Hubble Asteroid Hunter III. Physical properties of newly found asteroids
Hubble Asteroid Hunter III. Physical properties of newly found asteroidsHubble Asteroid Hunter III. Physical properties of newly found asteroids
Hubble Asteroid Hunter III. Physical properties of newly found asteroids
 
Pests of cotton_Borer_Pests_Binomics_Dr.UPR.pdf
Pests of cotton_Borer_Pests_Binomics_Dr.UPR.pdfPests of cotton_Borer_Pests_Binomics_Dr.UPR.pdf
Pests of cotton_Borer_Pests_Binomics_Dr.UPR.pdf
 
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 60009654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
 
Presentation Vikram Lander by Vedansh Gupta.pptx
Presentation Vikram Lander by Vedansh Gupta.pptxPresentation Vikram Lander by Vedansh Gupta.pptx
Presentation Vikram Lander by Vedansh Gupta.pptx
 
DIFFERENCE IN BACK CROSS AND TEST CROSS
DIFFERENCE IN  BACK CROSS AND TEST CROSSDIFFERENCE IN  BACK CROSS AND TEST CROSS
DIFFERENCE IN BACK CROSS AND TEST CROSS
 
GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)
 

DSD-INT 2014 - Data Science symposium - Application 2 - System identification with evolutionary computing, Dr. Christiaan Erdbrink, University of Amsterdam

  • 1. Christiaan Erdbrink chriserdbrink@gmail.com Data Science Symposium 31.10.2014 System identification using evolutionary computing
  • 2. My background MSc Delft University of Technology, Civil Engineering, fluid mechanics Deltares, flow around hydraulic structures PhD University of Amsterdam, Computational Science Saint Petersburg National Research University of Information Technologies, Mechanics and Optics (ITMO University)
  • 3. Christiaan Erdbrink Problem description Solution strategies - traditional - CFD - data-driven Evolutionary computing Conclusions & Outlook on future work Questions/Discussion Outline
  • 4. Christiaan Erdbrink nl.wikipedia.org infopuntveiligheid.nl coflexbouweninfra.noordhoff.nl Problem description
  • 5. Christiaan Erdbrink Haringvliet barrier Problem description
  • 6. Christiaan Erdbrink Physics of flow-induced vibrations (in a nutshell) Problem description Excitation mechanisms turbulence stable vortex shedding flow instabilities self-excitation unstable fluid resonance
  • 7. Christiaan Erdbrink Physics of flow-induced vibrations (in a nutshell) Problem description observed response excitation mechanisms assessment measures gate design real-life conditions
  • 8. Christiaan Erdbrink Problem description Solution strategies - traditional - CFD - data-driven Evolutionary computing Conclusions & Outlook on future work Questions/Discussion
  • 9. Christiaan Erdbrink Haringvliet barrier Traditional solutions
  • 10. Christiaan Erdbrink gate Traditional solutions Erdbrink, Krzhizhanovskaya, Sloot (2014): “Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids & Structures.
  • 12. Christiaan Erdbrink A/D = f( ζ , mr , Vr , Fr, St, I ) f a Δh Cs Traditional solutions Vr (-) Fz (-) Erdbrink, Krzhizhanovskaya, Sloot (2014): “Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids & Structures.
  • 13. Christiaan Erdbrink xi (t)  xi (t)  xi a Δh f Traditional solutions
  • 14. Christiaan Erdbrink Problem description Solution strategies - traditional - CFD - data-driven Evolutionary computing Conclusions & Outlook on future work Questions/Discussion
  • 15. Christiaan Erdbrink gate with leakage numerical simulations: CFD Erdbrink, Krzhizhanovskaya, Sloot (2014): “Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids and Structures.
  • 16. Christiaan Erdbrink at Vr ≈ 10: numerical simulations: CFD Erdbrink, Krzhizhanovskaya, Sloot (2014): “Reducing cross‐flow vibrations of underflow gates: experiments and numerical studies”, J of Fluids and Structures.
  • 17. Christiaan Erdbrink Problem description Solution strategies - traditional - CFD - data-driven Evolutionary computing Conclusions & Outlook on future work Questions/Discussion
  • 18. Christiaan Erdbrink h1(t) h2(t) safe or unsafe data-driven solution xi (t)  xi (t)  xi (t) a(t) f(t) xcrit
  • 19. Christiaan Erdbrink Use classification to avoid critical regions data-driven solution Vr (-) a (m) Erdbrink, Krzhizhanovskaya, Sloot (2012): “Controlling flow-induced vibrations of flood barrier gates with data-driven and finite-element modelling”, FLOODrisk2012
  • 20. Christiaan Erdbrink Designing a control system data-driven solution
  • 21. Christiaan Erdbrink Problem description Solution strategies - traditional - CFD - data-driven Evolutionary computing Outlook & Conclusions Questions/Discussion
  • 22. Christiaan Erdbrink approaches evolutionary computing traditional field measurements physical modelling FIV problems numerical simulations data-driven for control: for system id: classification evolutionary computing signal analysis FEM CFD, CFSI classic machine learning differential evolution genetic programming
  • 23. Christiaan Erdbrink evolutionary computing Hornby et al. (2006): “Automated antenna design with evolutionary algorithms” en.wikipedia.org/wiki/Evolved_antenna
  • 24. Christiaan Erdbrink Evolutionary algorithms Eiben & Smith (2011): “Introduction to evolutionary computing” evolutionary computing
  • 25. Christiaan Erdbrink Evolutionary Computing Meta-heuristics evolutionary computing
  • 26. Christiaan Erdbrink In general, EAs work well: - for multimodal problems - for multi-objective optimization - in hard design problems where a proposed configuration can be tested unambiguously - when small improvements are appreciated - when speed is not essential - when standard methods fail evolutionary computing
  • 27. Christiaan Erdbrink Reverse engineering dynamical systems For example, evolutionary computing Erdbrink, Krzhizhanovskaya: “Identifying Self‐Excited Vibrations with Evolutionary Computing”, Procedia Computer Science, Vol.29, pp.637‐647. Sensitivity analyses population size termination model parameters evaluation tolerance solver type
  • 28. Christiaan Erdbrink Fitness progression - model parameters updated once in 20 gens updated each gen evolutionary computing
  • 29. Christiaan Erdbrink 0 2 4 6 8 10 x 10 4 2 4 6 8 10 12 generation fitness best Fitness progression - termination criterion 0 500 1000 1500 2 4 6 8 10 12 generation fitness best evolutionary computing
  • 30. Christiaan Erdbrink Genetic Programming – applied to Symbolic Regression (x-C)*log(2x) y = (x-2.7139)*log(2x) *-LxC+xx evolutionary computing y = f(x,y) , etc.
  • 31. Christiaan Erdbrink Example y = 0.468e|x|sin(3.795x) evolutionary computing
  • 32. Christiaan Erdbrink evolutionary computing Mining physical systems (the “robot scientist”) M Schmidt, and H Lipson Science 2009;324:81-85
  • 33. Christiaan Erdbrink Application example evolutionary computing C.D. Erdbrink (2014): “Modelling flow-induced vibrations of gates in hydraulic structures”, PhD thesis Univ. of Amsterdam
  • 34. Christiaan Erdbrink Problem description Solution strategies - traditional - CFD - data-driven Evolutionary computing Outlook & Conclusions Questions/Discussion
  • 35. Christiaan Erdbrink Conclusions Data-driven methods should not be seen as competitors of traditional forms of modelling, but as valuable complementary tools. Monitoring of gate behaviour combined with classification of dynamic response states can be used to avoid critical vibration ranges. Evolutionary Computing… …is a versatile approach for all kinds of optimization problems. …has evolved from a hobby for computer scientists to an important area of research, with innumerous successful applications. …can be applied to output-only identification of (complex) dynamical systems. …is capable of automatically deriving meaningful elementary equations and from data.
  • 36. Christiaan Erdbrink chriserdbrink@gmail.com Thank you for your attention!