Академический Документы
Профессиональный Документы
Культура Документы
Control Engineering
8-1
Data
Identification
Model
White-box identification
estimate parameters of a physical model from data
Example: aircraft flight model
Gray-box identification
Rarely used in
real-life control
Black-box identification
determine model structure and estimate parameters from data
Example: security pricing models for stock market
Control Engineering
8-2
Aerospace
white-box identification, specially designed programs of tests
Automotive
white-box, significant effort on model development and calibration
Disk drives
used to do thorough identification, shorter cycle time
Embedded systems
simplified models, short cycle time
EE392m - Winter 2003
Control Engineering
8-3
0.8
0.6
0.4
0.2
0
TIME
-0.5
Control Engineering
TIME
8-4
Actuator bumped
0.5
0
0
200
400
600
TIME (S EC)
800
1000
100
Control Engineering
200
300
400
TIME (S EC)
500
600
8-5
Noise reduction
Noise can be reduced by statistical averaging:
Collect data for mutiple steps and do more averaging to
estimate the step/pulse response
Use a parametric model of the system and estimate a few
model parameters describing the response: dead time, rise
time, gain
Do both in a sequence
done in real process control ID packages
Pre-filter data
Control Engineering
8-6
Linear regression
Mathematical aside
linear regression is one of the main System ID tools
Data
Regression weights
Regressor
y (t ) = j j (t ) + e(t )
Error
y = + e
j =1
y (1)
1 (1) K K (1)
1
e(1)
O
M , = M , e = M
y = M , = M
y ( N )
1 ( N ) K K ( N )
K
e( N )
Control Engineering
8-7
Linear regression
Makes sense only when matrix is
tall, N > K, more data available than
the number of unknown parameters.
Statistical averaging
||e||2
min
y = + e
= ( T ) T y
1
Correlation interpretation:
N
N 2
K K (t )1 (t )
1 (t )
t =1
1 t =1
,
M
O
M
R=
N
N N
2
1 (t ) K (t ) K
K (t )
t =1
t =1
EE392m - Winter 2003
Control Engineering
= R 1c
N
1 ( t ) y ( t )
1 t =1
M
c=
N N
K (t ) y (t )
t =1
8-8
1 T
T
= ( ) y
Main issue:
small error in a might mean large change in response
EE392m - Winter 2003
Control Engineering
8-9
Regularization
Linear regression, where T is ill-conditioned
Instead of ||e||2 min solve a regularized problem
2
2
e + r min
y = + e
r is a small regularization parameter
Regularized solution
1 T
T
= ( + rI ) y
Control Engineering
8-10
Regularization
Analysis through SVD (singular value decomposition)
= USV T ;
V R n ,n ;U R m ,m ; S = diag{s j }nj =1
Regularized solution
n
= ( T + rI ) T y = V diag 2 j U T y
s j + r j =1
s
s 2 + 0.1
1
0.5
0
Control Engineering
8-11
y ( t ) = h ( k )u ( t k ) + e( t )
k =1
10
20
30
40
PRBS =
Pseudo-Random Binary Sequence,
see IDINPUT in Matlab
1 (t ) = u(t 1)
M
K (t ) = u (t K )
EE392m - Winter 2003
h (1)
= M
h ( K )
Control Engineering
1 T
T
= ( + rI ) y
8-12
50
PRBS excitation
input
1
0.5
0
-0.5
-1
0
200
400
600
800
1000
800
1000
Simulated system 1
output: 4000
0.5
samples, random 0
-0.5
noise of the
-1
amplitude 0.5
0
200
400
600
TIME
Control Engineering
8-13
Impuls e Re s pons e
Impulse response
for the simulated
system:
0.2
0.15
0.1
0.05
0
-0.05
T=tf([1 .5],[1 1.1 1]);
0
P=c2d(T,0.25);
Time (s ec)
Control Engineering
8-14
Loss Index V
V = y (t ) y (t | )
Loss index
J = y (t ) y (t | )
Optimizer
y (t )
u (t )
Control Engineering
8-15
g
=
TD
g
TD
Example:
Paper
machine
process
Control Engineering
8-16
Gain estimation
For given , TD , the modeled step response can be
presented in the form
y (t | ) = + g y1 (t | , TD )
This is a linear regression
2
w1 = g
k =1
w2 =
y (t | ) = wk k (t )
1 (t ) = y1 (t | , TD )
2 (t ) = 1
y (t | , TD ) = + g y1 (t | , TD )
Control Engineering
8-17
V = y (t ) y (t | , TD )
t =1
Control Engineering
8-18
1.6
1.4
Nonlinear
Regression ID
0.8
0.6
1.2
1
0.4
0.8
Linear
Regression ID
of the first-order
model
0.2
0
-0.2
-0.4
10
20
30
40
50
60
70
0.6
Nonlinear
Regression ID
0.4
0.2
0
80
-0.2
0
100
200
300
400
500
600
700
time in s e c.; MD re s pons e - s olid; e s tima te d re s pons e - da s he d
Control Engineering
8-19
800
Linear filtering
A trick that helps: pre-filter data
Consider data model
y = h *u + e
L( h * u ) = ( Lh ) * u = h * ( Lu )
Control Engineering
8-20
Multivariable ID
Apply SISO ID to various input/output pairs
Need n tests - excite each input in turn
Step/pulse response identification is a key part of the
industrial Multivariable Predictive Control packages.
Control Engineering
8-21