Professional Documents
Culture Documents
Elevator
M. Moazen1, M.B.B.Sharifian2
H. Afshari3
1, 2
Fig. 1.
Fig. 2.
I.
INTRODUCTION
m(t ) F (t ) F (i, ) F (t )
K
d
ki 2 (t )
F (i, ) 2
(t )
2k
2ki (t )
i (t ) 2 (t )
u (t ) ri(t )
(
t
)
(t )
m(t ) FK (t ) F (i, ) Fd (t )
(1)
2
0 A
F (i, )
0 AN 2 i 2 (t )
4 2 (t )
(3)
where N and i(t) are the number of winding turns and the
current of single Maglev equipment, respectively. Voltage of
single Maglev equipment is attained by:
u (t ) ri(t )
0 AN 2
AN 2 i(t )
i (t ) 0 2
(t )
2 (t )
2 (t )
(4)
0 AN 2
4
x2
x1
0
0
2
k x 3
F
(
t
)
F
(
t
)
d
K
x 2 0 u
m
m x1 x1
m
x
0
3 x 2 x3 rx1 x3
2k
x 2k
1
y x
1
where
(6)
F (t )
FK (t )
can be seen as the disturbance as d
.
m
m
III.
(2)
where 0 and A are permeability of the vacuum and crosssection of the air gap, respectively.
by:
where k
(5)
CONTROL STRATEGY
A. Feedback Linearization
If the state variable vector is chosen as follows:
z z1
z2
z3
x1
x2
k x
3
m x1
(7)
z 0
y 1
1 0 0
0 1 z 0 v Az Bv
0 0 1
0 0 Cz
(8)
rx32
x
3 u . Therefore nonlinear voltage
mx1 mx1
feedback control principle is given by:
where v
u ( x) ( x)v rx3
mx1
v
x3
(9)
ref
PID
e(t ) yexp y
Eq. 9
Fd
Electromagnet
z3
i
FK
(10)
(11)
(12)
z A 0 z B
0
v
p C 0 p 0
y exp
y C 0 p
(13)
de(t )
v kz k p e(t ) k i e(t ) dt k d
0
dt
Fig. 3.
p (t ) e(t ) yexp Cz
d/dt
State Feedback
(14)
k k 0
ref 0.003m
N 400turns
m 500kg
A 0.65 0.65m 2 .
The gains of the PID controller are found by trial and error
method for best response for the test system.
A. Test I
In this section, the Maglev system using the described
and
feedback
linearization
method
(for
0 0
FK (t ) 5000 N ) has been tested. The system model in
MATLAB software is seen as Fig. 3. The simulation results are
shown in Figs. 4 to 9. Each air gap changing factor could be
considered as a disturbance force.
coefficients.
delta [mm]
IV.
SIMULATION
and
Fig. 4.
deltaref
delta
0.5
1
t [s]
1.5
600
400
f f[N/m]
d [N]
200
where
0
-200
-400
-600
Fig. 5.
(15)
0.5
1
t [s]
1.5
F (i1 , )
0 AN 2 i12 (t )
4 2 (t )
(16)
F (i2 , )
10000
8000
(17)
[N]
[N/m]
FF
6000
4000
2000
0
Fig. 6.
0 AN 2 i22 (t )
.
2
42 ref (t )
0.5
1
t [s]
1.5
V [mm/s]
200
0
-100
Fig. 7.
100
0.5
1
t [s]
1.5
u [V]
1000
0
-1000
-2000
Fig. 8.
0.5
1
t [s]
1.5
i [A]
0
-1
-2
-3
Fig. 9.
0.5
1
t [s]
1.5
V.
(18)
A. Test II
The system model in this case in MATLAB software is
seen as Fig. 10. In this case, because the current and thus
response of Maglev force is directly controlled, the system has
greater stability and won't need to state feedback and
consequently, the control system will be simple.
The simulation results are shown in Figs. 11 to 16
200
PID
Fd
not
>=0
Fig. 10.
i1
Electromagnets
i2
100
V [mm/s]
ref
0
-100
Fig. 15.
0.5
1
t [s]
1.5
deltaref
delta
i1
i2
i [A]
delta [mm]
1
0
Fig. 11.
0.5
1
t [s]
1.5
-1
Fig. 16.
0.5
1
t [s]
1.5
600
delta [mm]
0
-200
-400
-600
Fig. 12.
deltaref
delta
200
[N]
f fd[N/m]
400
0.5
1
t [s]
1.5
3.5
3
2.5
Fig. 17.
0.5
1
t [s]
1.5
F1
F2
400
5000
200
0
[N]
f fd[N/m]
FF[N/m]
[N]
10000
Fig. 13.
-200
-400
0.5
1
t [s]
1.5
-600
Fig. 18.
0.5
1
t [s]
1.5
200
F1
F2
600
100
[N]
[N/m]
FF
[N]
F F[N/m]
800
F1
F2
150
50
400
200
Fig. 14.
0.5
1
t [s]
1.5
Fig. 19.
0.5
1
t [s]
1.5
V [mm/s]
10
0
-10
0
Fig. 20.
0.5
1
t [s]
1.5
REFERENCES
[1]
[2]
i [A]
i1
i2
0
-0.5
Fig. 21.
[3]
0.5
1
t [s]
1.5
VI.
[4]
[5]
CONCLUSION
[6]
[7]
[8]