power systems control prof. wonhee kim...2 stability in state space equation: state feedback u u t t...
TRANSCRIPT
Power Systems Control
Prof. Wonhee Kim
Ch.3. Controller Design in Time Domain
2
Stability in State Space Equation: State Feedback
x x u
y x u
t A t B t
t C t D t
u xt K t
x x x
x
t A t BK t
A BK t
K should be chosen such that (A-BK) is Hurwitz!
3
Observer DesignSystem
Observer design
- Aim:
- Estimation error define:
- Observer
- Estimation error dynamics
x Ax Bu
y Cx
x̂ x
ˆx x x
ˆ ˆ ˆ
ˆ ˆ
x Ax Bu L y y
y Cx
ˆ ˆ ˆx x x Ax Bu Ax Bu LC x x
A LC x
L should be chosen such that (A-LC) is Hurwitz!
System
Control input using estimated state
System with control input
Closed-loop System
4
Separation Principle
x Ax Bu
y Cx
ˆu Kx
ˆx Ax BKx
A BK x BKx
x A BK x BKx
x A LC x
K and L can be separatively designed
5
Controllability and Observability
6
Controllability and ObservabilityControllability: for any desired eigenvalues, K exists such that (A-BK) has the desired
eigenvalues
Controllability matrix has n rank.
Observability: for any desired eigenvalues, L exists such that (A-LC) has the desired
eigenvalues
Controllability matrix has n rank.
1n
conC B BA BA
1n
T T T T T T
obO C A C A C
1
ob
n
C
CAO
CA
7
State Feedback Controller Design for Reference Tracking
Regulation)
Using final value theorem, steady-state response with step reference should be 1.
where Kdc is a dc gain for C(sI-(A-BK))-1B
x x u
y x
t A t B t
t C t
u xt K t Jr t
x x x
x
t A t BK t BKr t
A BK t BJr t
x x u
y x
t A t B t
t C t
1Y sT s C sI A B
U s
1Y sT s C sI A BK B J
R s
0 0 0
lim lim lim 1dcs s s
y sT s R s T s JK
1
dc
JK
8
Tracking Controller Design1. System
2. Controller design
- Aim:
- Tracking error define:
- Tracking Error dynamics
- Controller
1 2
2 3
1 1n n n
x x
x x
x a x a x bu
1x r
1 1
1n
n n
e r x
e r x
1 1 2
1 1
n n
n n n n
e r x e
e r x r a x a x bu
1 2
1 1n n n
e e
e k e k e
1 1 2 2 1 1
1 n
n n n nu r a x a x a x k e k eb
9
Tracking Controller Design Example1. System
2. Controller design
- Aim:
- Tracking error define:
- Tracking Error dynamics
- Controller
Control gain k1, k2 should be designed such that Ac = [0 1; k1 k2] is Hurwitz
1 2
2 1 1 2 2
x x
x a x a x bu
1x r
1 1
2 2
e r x
e r x
1 1 2
2 2 1 1 2 2
e r x e
e r x r a x a x bu
1 2
2 1 1 2 2
e e
e k e k e
1 1 2 2 1 1 2 2
1u r a x a x k e k e
b
10
Tracking Controller II Design1. System
2. Desired state dynamics: ud is designed for the desired state
- Tracking error:
- Aim:
- Tracking error dynamics:
- Controller
x Ax Bu
y Cx
0 asde x x t
d
d d
d
e x x
Ax Bu Ax Bu
Ae Bu Bu
du u Ke
d d d
d d
x Ax Bu
y Cx
de x x
d d
d d
e Ae Bu B u Ke
Ae Bu Bu BKe
A BK e
11
Tracking Controller II Design: DC Motor1. System
2. Desired constant velocity ωd, id and vd:
3. Desired state dynamics: ud is designed for the desired state
- Tracking error:
- Aim:
- Tracking error dynamics:
- Controller:
1
1t
B KiJ
i K Ri vL
0 asde x x t
d
d d
d
e x x
Ax Bu Ax Bu
Ae Bu Bu
1 2
d
t d d d d
u u Ke
K Ri k k i i
d d d
d d
x Ax Bu
y Cx
de x x
d d
d d
e Ae Bu B u Ke
Ae Bu Bu BKe
A BK e
1 1
0 , 0d d t d d dB Ki K Ri vJ L
,d dd d t d d t d
B BRi v K Ri K
K K
12
Control Algorithm Implementation
DC Motor System
Velocity feedback (y=ω)
Input voltage
(v)
1
1t
B KiJ
i K Ri vL
y
1
2
1 ˆˆ ˆ ˆ
1ˆ ˆˆ ˆ
ˆˆ
t
B Ki l y yJ
i K Ri v l y yL
y
1 2ˆˆ
dd
t d d d d
Bi
K
v K Ri k k i i
Control Board
d
vˆˆ , i
13
Control Algorithm Implementation
DC Motor SystemControl Board
DC Motor
+ Encoder
Input voltage
(v)
Velocity feedback (y=ω)
14
Control Algorithm Implementation
DC Motor SystemControl Board
DC Motor
+ Encoder
Input voltage
(v)
Velocity feedback (y=ω)
Load
disturbance
Load disturbance may degrade control performance of DC motor!
15
Tracking Controller: Position Control of DC Motor
DC motor with unknown load disturbance, d for position control:
Not normal form!
Definition of Acceleration state variable
With α, DC motor model:
Now DC motor model becomes normal form, thus, tracking controller can be applied!
1
1t
B Ki dJ
i K Ri vL
2
2 2 2
1
1
t
B Ki dJ
B KiJ
KKB BK B KR Ki d i v
J J J JL JL JL
Note that the disturbance can include
the parameter uncertainties as well as
the external disturbance
16
Extended State Observer Design: DC MotorDC motor with unknown load disturbance, d for position control
Extended state d with assumption that d is constant
Extended state observer to estimate full state and disturbance using position feedback
1
1t
B Ki dJ
i K Ri vL
1
1
0
t
B Ki dJ
i K Ri vL
d
1
2
3
4
ˆ ˆˆ
1 ˆ ˆˆˆ ˆ
1 ˆˆ ˆˆ
ˆ ˆ
t
l
B Ki d lJ
i K Ri v lL
d l
17
Extended State Observer Design: DC MotorEstimation error
Estimation error dynamics
ˆˆ ˆˆ, , ,i i i d d d
1
2
3
4
1 ˆˆˆ
1 ˆˆt
l
B Ki d lJ
i K Ri lL
d l
1
2
3
4
1 0 0
1
0
0 0 0
t
A
l
B Kl
J J J
K iRi lL L d
d l
l1, l2, l3 and l4 should be chosen such that A is Hurwitz!
18
Experimental Results
Position control with normal observer
(without disturbance rejection)
Position control with ESO
(with disturbance rejection)