Professional Documents
Culture Documents
Tristan Perez and Mogens Blanke
Tristan Perez and Mogens Blanke
Motivation
The effects of roll on ship performance became more
noticeable in the mid-19th century when:
W. Froude (1819-1879)
Outline
Working Principles,
Performance,
Historical aspects
Models,
Wave-induced motion
Simulation and control design models
Objectives,
Fundamental limitations,
Control strategies
Research Outlook
Bilge Keels
U-tanks
Gyro-stabilisers
1914 to 1925
active control
1914,1924
Twin Gyros
Schlick
(Germany)
Sperry
(USA)
Fieux
(France)
Today
Time
Halcyon
Ferretti
Sea Gyro
Shipdynamics
Seakeeper
Fin-stabilisers
60% to 90% roll reduction (RMS)
Performance depends on speed
Control is important for performance
Easy to damage
Most Expensive stabiliser
Can produce noise affecting sonar
Dynamic stall
Rolls-Royce
10
Historical Aspects
11
! n "
pb/n
= [N, E, D, , , ]T .
!
13
nb
pb/n
bb/n
"
= [u, v, w, p, q, r]T .
Kinematic model:
= J()
Kinetic model:
1DOF, 4DOF
Force to motion
14
1DOF: roll
4DOF: surge, roll, sway, and yaw
= p,
Kinematic model:
Ixx p = Kh + Kc + Kd
Kinetic model:
Hydrodynamic
moments
Control
moments
Disturbance
moments
Hydrodynamic Moments:
Viscous
Effects
Hydrostatic
Effects
= [ ] ,
T
= [u v p r]T ,
i = [Xi Yi Ki Ni ] ,
T
Kinematic model:
Kinetic model:
= p,
= r cos r
MRB + CRB () = h + c + d
h MA CA () D() K()
16
Ocean Environment
Usual assumptions for sea surface elevation
Zero-mean
(t):
(, )
17
Speed
Encounter angle
2 U
e =
cos()
g
18
(e ) = !!
!1
()
2U
g
!
!
cos()!
S ()
Wave spectrum
F(j,,U)
G(j,U)
G11 (j, U )
..
G(j, U ) =
.
G61 (j, U )
Roll RAO:
Motion Time
series
G16 (j)
..
2
1
=
([M
+
A()]
+
jB()
+
G)
RB
.
G66 (j, U )
H4 (j, , U ) =
6
!
k=1
19
S ()
6
!
73
k=1
|H4 (, U, )|
&!*+,
'#*+,
(!*+,
-#*+,
.!*+,
$!#*+,
$%!*+,
$&#*+,
$#!*+,
$(#*+,
(
'
%
&
!
%!!
%"#
%
$#!
$"#
$!!
$
#!
!"#
!
Fig. 4.6. motion RAO Roll of the naval vessel benchmark at 15 kt for different
angles.
RAO
are given
as
a function of the wave frequency.
20 encounter
8th IFAC
CAMS,The
Septmotion
15th-17th,
Rostock,
Germany
Actuators
u
Output sensitivity
cl (s)
S(s) !
ol (s)
22
Kc
ol
Ship
cl
Control
Roll spectrum relations
cl () = |S(j)|2 ol ()
log |S(j)| d = 0
cl () = |S(j)|2 ol ()
24
q = q + j0
q
W (q, ) = 2
q + 2
25
Example Blanke
et al (2000)
'())*+,-).*'./0123(,
!&"
26
"
#!
#"
4+5.6.'3(/*78.19
$!
$"
%!
(t)
27
2rudder
b
4rudder zCP
L
(8)
6rudder xbCP L,
b
are the coordinates of the cenwhere xbCP and zCP
ter of pressure of the rudder (CP ) with respect to
IDOF model including fin forces:the b-frame. The center of pressure is assumed to
be located at the rudder stock and half the rudder
span. The rudder data
is shown in Table A.4.
= p,
[Ixx + Kp ] p + (Kp +
Control issues:
Parametric uncertainty
Sensitivity-integral constraints
Control strategies:
=
K
2U
Kfin.
PID, Hinf
28
ing
the
is a
The
with
and
bas
(Pe
for
CG
ob
zb
yb
rbCP
CP
y!
y !!
z !! z !
in
whe
and
%
n
For
CL [-]
CL [-]
e [deg]
[deg]
e [deg]
Time [s]
30
Fin Stabilisers
Perez & Goodwin (2003): MPC constraints effective angle of attack.
31
Cowley & Lambert (1972) used roll fbk, reported yaw interference.
32
!S(j)!
33
1
1
"
(1 ,2 )
(1 ,2 )
J = E[2 + (1 )( d )2 ]
Minimum variance case (=1):
34
E[ ] 2 q (q)
2
Maximum attained
here
35
35
36
37
38
Performance Limitations
39
40
41
Control Strategies
PID,
Hinf
Loop
shaping
Adaptive LQG
Model Predictive Control
Switched control
Nonlinear control
42
42
c (1 + sz1 )(1 qs )
(1 + sp1 )(1 +
s2
sp2 )( 2
p
*+,-./01
+ 2p sp + 1)
;+,-./01
&'(
!&'(
&'%#
!&'(:
&'%!
Changes in dynamic
response due to speed
coupled with changes in
disturbance spectrum
results in a strong need
for adaptation.
&'%)
!&'9
&'%(
!&'9:
&'%
!
"
#
$
%&
23//4-56728
%%
"
;+,-31</%
#
$
%&
23//4-56728
%%
;+,-31</(
!&'9
!&'&(:
!&'9:
!&'&9
!&')
!&'&9:
!&'):
!&'&)
!&':
!&'&):
!&'::
!
43
"
#
$
%&
23//4-56728
%%
"
#
$
%&
23//4-56728
%%
H2 Optimal Design
() = Hd (j)Hd (j)
1
(G )
s+q
= G (s)
sq
V = Hd G
1
Hd Hd
sq
1
s+q Hd Hd
s+q
Hd (s)
= Hd (s) + Hd+ (s)
sq
This shows the strong dependency on sea state, sailing conditions and
changes in the vessel model --- adaptation is required.
44
Sd (s) =
s2
d2
(1 +
+ 2d sd + 1
s
d )(1
s
d )
s
C
(s) = (Sd (s)1 1)G1
(s)
45
1+
s
q
s
q
P (s)1
k1 s
=
c d
s2
p2
s2
d2
+ 2p sp + 1 (1 + s )(1 + s )
p1
p2
s
+ 2d s + 1 (1 + sz1 )(1 + q )
d
::;*4+5.*8.,8323532<*=>?:@*!*AB%CDE+5FD$G#!
#&"
Successful performance
was demonstrated for the
SF300 patrol boat of the
Danish Navy
'())*+,-).*'./0123(,
!&"
46
"
#!
#"
4+5.6.'3(/*78.19
$!
$"
%!
Gyro Stabilisation
!s
!m
Kg = Is s
!p
!g
!w ,
l
n
I44 + B44
+ B44
|| + C44 = w Kg cos ,
! "# $
g
Ig
+ Bg + Cg sin = Kg cos +p
! "# $
m
47
Gyro as a Damper
p : q
e
n- number of gyros
q-constant
48
Gyro-control Design
(s)
(s)
Kg s
G(s) =
=
=
,
2
!
!
(s)
Ig s + Bg s + Cg
(s)
Bg!
= Bg + Kr ,
Cg!
= Cg + Ka
p : q
|G()| q,
arg G() 0
Ship
gyro
This can be achieved by forcing two real poles on G(s) (Perez Steinman, 2009)
49
50
50
IV - Research Outlook
New Devices
Unsolved Problems
New Devices
Flapping fins (Fang et al. 2009 Ocean Engineering 36)
Added mass
52
New Devices
Rotor Stabiliser (Quantumhydraulic.com)
53
Parametric roll
54
Conditions for PR
The following conditions can trigger the effect:
K() = g GZ()
56
Videos
57
Conclusion
For over 100 years different devices have been proposed to control
roll motion (reduction)
58
Fundamental limitations
Gyros, Flap fins, rotor stabilisers (need for zero speed performance)
Research outlook
59
Thank You