system identification in dynamic networks · system identification g + v u y •use of models, for...
TRANSCRIPT
System Identification:from open-loop and closed-loop systems
to dynamic networks
Paul M.J. Van den Hof
Inauguration speech,
Hungarian Academy of Sciences, October 2, 2017, Budapest, Hungary
System identification
• Estimating dynamic models on the
basis of noisy observations over time
G +
v
u y
• Dynamic cause-effect relationships described by
(linear) ordinary differential equations (ODE’s)
• Models in physical structure or “black-box”
System identification
G +
v
u y
• Omnipresent/ubiquitous in many branches of
science and technology
• from econometrics to mechatronics, and
• from medicine / disease treatment to robotics
• from automotive systems to industrial processes
System identification
G +
v
u y
• Use of models, for
• understanding
• prediction of future behaviour
• simulation
• indirect measurement / monitoring
• control
• Model-based control and optimization is the leading
paradigm for the optimal operation of dynamical
systems
• Model = knowledge base,
stored structured information from the past
+ G +
v
C- u yr
A “man-on-the-moon” example
• Moving mass 90 kg
• Scan a wafer in 13 sec,
0.13 sec per chip.
• Overlay of around 2 nm.
https://youtu.be/jH6Urfqt_d4?t=2m29s
ASML Wafer scanner
System identification – the birth
Proc. IFAC Symp. Self-Adaptive Control Systems,
Teddington, England, 1965Karl J. Åström Torsten Bohlin
Rudolf KalmanEffective construction of linear state-variable models from
input-output functions.
Regelungstechnik, Vol. 14 (1966), 545-548.
System identification
1974Pieter Eykhoff
Goodwin and Payne (1977)Graham Goodwin
System identification – Prediction Error Approach
Lennart Ljung
(1987) (1999)
System identification
Rik Pintelon and Johan Schoukens
(2000,2012), Frequency domain approach
My own entrance into the field
Tibor Vamos Janos Gertler Laszlo Keviczky
Contents
• Introduction
• Developments and some highlights
• From SISO to MIMO
• Orthogonal basis funcitons
• Closed-loop identification
• Identification for control
• The role in future technology
• Towards dynamic network identification
Developments and some highlights
• From SISO to MIMO (multi-input multi-output)
Gu y
Just technicalities and added complexity?
How to parsemoneously and uniquely parametrize models through
a real-valued vector to be estimated from data?
For number of outputs p > 1 this is a nontrivial problem
For black-box models:
From SISO to MIMO
Options:
State space models:
For p>1, there exists no continuous parametrization that covers all models,
with a predefined order (state space dimension) [Hazewinkel and Kalman, 1976]
Matrix fractions / Vector difference equations:
with matrix polynomials in the forward and backward shift operators
Focus of many contributions in the 1980s, on the construction of canonical and
pseudo-canonical (overlapping) forms.
From SISO to MIMO
Focus of many contributions in the 1980s, among which Gevers, Wertz, Guidorzi,
Bokor and Keviczky, Correa and Glover, Rissanen, ....
After 1990s in identification
• succeeded by subspace and other techniques,
• less sensitive to lack of uniqueness of the parameters – models mapping
Important insights into the structural properties of linear MIMO systems
Set of all models with order n
Orthogonal basis functions (OBFs)
0 10 20 30 40-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
Pulse response representation:
As a parametrized model set:
with complete set of orthogonal basis functions
of the function space (all stable models)
with possibly large values of
Are there appropriate generalizations of the set of basis functions
that allow a fast convergence of the series?
Orthogonal basis functions (OBFs)
0 10 20 30 40-0.2
0
0.2
0.4
0.6
0.8
Consider
where the functions have dynamics
(prior info)
Examples:
Laguerre:
Orthogonalized version of
Takenake/Malmquist:
Orthogonalized version of
GOBF’s: use repeated set of poles
Orthogonal basis functions (OBFs)
Generalization of tapped delay line:
With all-pass functions
Analysis and completeness:
Built on the clasic work of a.o. Otto Sász (1955) and Gábor Szegő (1939,1958)
Orthogonal basis functions (OBFs)
GOBFs: very effective model structure for identification:
• identification properties fully analyzed (VdH, Heuberger, Bokor, 1995)
• simple (linear regression) algorithms
• appropriate handling of prior info
• fit for large scale problems
• used for uncertainty quantification
• for nonlinear (LPV) modeling
• and in system (realization) theory
Orthogonal basis functions (OBFs)
GOBFs: very effective model structure for identification:
Hambo Transform
Closed-loop identification
+ G +
v
C
- u yr
• Many systems operate under the
presence of feedback (control)
• Intrinsic problem that input and
disturbances are correlated
• Plant input u is only partly under control of
a designed experiment rr, u and y are measured
Classical approach:
• Direct parametric method, based on u,y, (Swedish school, 1970s)
• Consistent and minimum variance (CRLB), under perfect conditions
No well-interpretable results in case of approximations
Closed-loop identification
Plant representation
white noise
and uncorrelated
+ G0 +
v
C
- u yr
H0
eProjection/two-stage/IV method
Use measured excitation signal ,
to generate ,
the signal projected onto
[Van den Hof & Schrama, 1993]
Estimate the dynamics between and
Closed-loop identification
Plant representation
white noise
and uncorrelated
+ G0 +
v
C
- u yr
H0
e1. Direct method
Consistent estimate of
provided that u is sufficiently exciting
2. Projection/two-stage/IV method
Consistent estimate of
provided that is sufficiently exciting
[Van den Hof & Schrama, 1993]
[Ljung, 1987]
Estimate achieves minimum variance
(CRLB)
No minimum variance properties
Implicit approximation properties
Tunable approximation properties
Identification for control
Feedback control systemIdentification
processprocessoutputinput
disturbance
1 0- 2
1 0- 1
1 00
1 0- 4
1 0- 2
1 00
1 02
10-2
10-1
100
-600
-400
-200
0
amplitude
fase
frequency
Data Model
Model Controller
Feedback control systemFeedback control system
controllercontroller processprocess+
-
outputreferenceinput
disturbance
Model-based control
Identification for control
controllercontroller processprocess+
-
outputreferenceinput
disturbance
• High interest from control community in 1990sGevers, Bitmead, Anderson, Schrama, de Callafon, Keviczky, Banyasz, Lee,
Skelton, Kosut, Hansen, Mareels, Bombois
• Primal result:
Optimal models for control are identified under
closed-loop conditions, with the to-be-designed
controller in the loop
Identification for control
ImplementatieImplementatieImplementation
RegelaarontwerpControl design
controller
IdentificatieIdentification
model
Experiment
data
ExperimentExperiment
evaluation
experiment design
Modelling for control is
learning(Gevers, 1993; Schrama, 1992;
Lee, Anderson et al, 1992,1993)
Solution through iterative
procedure
Modern version of
adaptive control
Identification for control
• In process control, experiments are typically expensive
Reaching the highest performance for an
expensive experiment
Designing the most economic experiment that
reaches the performance requirements
Least-costly experiment design (Bombois et al., 2006)
EU-STREP 2010-2013 – 2.5M€ www.fp7-autoprofit.eu
Goal of the project:
Advanced Autonomous Model-Based
Operation of Industrial Process Systems
Improved lifetime performance of model-based (MPC)
controllers by autonomous cost efficient maintenance.
Identification for control - Autoprofit
Autonomous maintenance for linear
model-based operation
Developments focused on:
• Performance monitoring and
diagnostics
• Autonomous testing:
• Autonomous MPC tuning:
• Extension to non-linear systems
• LPV modeling
Prototype testing on industrial
FT depropanizor (Sasol)
eco
no
mic
cri
teri
on
ba
se
d d
ecis
ion
s
least-costly
re-identification
Model structure
assessment
Retuning of
model-based
application
Performance
monitoring & diagnosis
selection of appropriate action
Degradation detected
Identification for control - Autoprofit
Contents
• Introduction
• Developments and some highlights
• From SISO to MIMO
• Orthogonal basis funcitons
• Closed-loop identification
• Identification for control
• The role in future technology
• Towards dynamic network identification
Industry 4.0 – process operations aspects
• integrated chains/networks of production units,
• fully automated, high level of sensing/actuation,
• data and product flows across classical
(company) borders (suppliers,customers,
energy grid)
• modular build-up
• continuously monitored for control,
optimization, (predictive) maintenance,
analysis, ......
• adapting to changing circumstances (process
and market conditons), and learning
• economically optimized
• supervised by new-generation HMI technology
and operators
From isolated (statically) optimized units to
[Boston Consulting Group report: “Industry
4.0, The Future of Production & Growth in
Manufacturing Industries“, 2015]
Cyber-physical systems of systems
Proposal of a European Research and Innovation Agenda on Cyber-Physical Systems of Systems, 2016-2025
Similar developments in other domains of technology
Role in future technology
Future requirements on engineering systems:
Handling of highly complex
interacting
distributed systems that
operate autonomously
with variable objectives
in a ``learning’’ mode
adapting to changing circumstances
and maintain a verifiable high performance
Role in future technology
Some required capabilities of models
1. Accuracy assessmenton-line assessment of model validity
2. Adaptabilityflexible on-line updating of models (dynamics and
interconnection structure)
3. Active data-driven learningdemands on accuracy, autonomy, robustness
active probing for information
all relating to phenomena of data-driven modeling
Data-driven modeling becomes an integral part in
virtually all complex engineering systems
Information-driven operations
From model-based control to information-driven operations
Models
Process data
Optimal proces operations
First principles
• Monitoring
• Indirect measurements
• Control
• (Predictive) maintenance
• Optimization
• Scheduling
Networked and distributed systems:
• collection of dynamical subsystems
• local control capability,
• physically interacting
Examples: smart power grids, intelligent traffic
networks, sensor networks & process plants
and their supply chain.
From single loops to interconnected systems
Contents
• Introduction
• Developments and some highlights
• From SISO to MIMO
• Orthogonal basis funcitons
• Closed-loop identification
• Identification for control
• The role in future technology
• Towards dynamic network identification
Dynamic network identification
C1
C2
G1
G2
G21
+
+
+
+ +
++
-
-
y1
y2
r1
r2
u1
u2
v1
v2
G10
G12
Gudi, R. D. and Rawlings, J. B. (2006). Identification for decentralized model predictive control.
AIChE Journal, 52(6):2198-2210.
Example decentralized MPC; 2 interconnected MPC loops
Target:
Identify interaction dynamics
Addressed by
Gudi & Rawlings (2006)
for the situation
(no cycles)
Dynamic network identification
C1
C2
G1
G2
G21
+
+
+
+ +
++
-
-
y1
y2
r1
r2
u1
u2
v1
v2
G10
G12
Example decentralized MPC; 2 interconnected MPC loops
Target:
Identify interaction dynamics
Structural aspects
need to be addressed.
Dynamic network identification
Some modules may be known (e.g. controllers)
ri external excitation
vi process noise
wi node signal
Dynamic network identification
Some modules may be known (e.g. controllers)
ri external excitation
vi process noise
wi node signal
Dynamic network identification
Some modules may be known (e.g. controllers)
ri external excitation
vi process noise
wi node signal
Dynamic network identification
Some modules may be known (e.g. controllers)
ri external excitation
vi process noise
wi node signal
Dynamic network identification
Relevant identification questions that appear:
How to perform local identification?
which signals to measure?
Dynamic network identification
Relevant identification questions that appear:
Where to optimally locate sensors and actuators?
Dynamic network identification
Relevant identification questions that appear:
How to identify a subnetwork?
which signals to measure?
Dynamic network identification
Relevant identification questions that appear:
How can we benefit from known (orange) modules?
Dynamic network identification
Relevant identification questions that appear:
Can we (on-line) identify the topology?
(notion of identifiability)
Dynamic network identification
Relevant identification questions that appear:
• Fault detection and isolation?
• Detect/identify/handle nonlinear elements
Dynamic network identification
• Attractive and rich domain
of research
• Many challenges ahead
• Relation to distributed
control / optimization
(multi-agent systems)
• Relate to developments in
machine learning / data
analytics
G1
C1
G2
C2
Gr
Cr
An exciting area with lots of new questions to be explored
Acknowledgements
• My teachers and supervisors:
Ad Damen, Pieter Eykhoff, Okko Bosgra and Jan Willems
• My co-authors, PhD students and postdocs, and
colleagues from TU Delft and TU Eindhoven
• My Hungarian colleagues and co-authors from SZTAKI,
Jozsef Bokor, Laszlo Keviczky, Szoltan Szabo
• International colleagues in the control community
• My sponsors, among which EU (FP7, ERC).
Acknowledgements
Köszönöm nagyrabecsülésüket és a megtiszteltetést, amit e rangos cím jelent számomra!
THE END