Beruflich Dokumente
Kultur Dokumente
a_greig@meng.ucl.ac.uk
r_moaleji@meng.uclac.uk
Abstract: Roll stabilization of ships at slow speeds or at rest has for long been a challenge
in Marine engineering. One of the common methods of doing so is using anti-roll tanks
which are available in passive form or an active form where a pump is used to move the
water. This paper presents a new strategy in controlling the actuating pumps of an active
U-tank with an adaptive inverse controller using filtered-x least mean square algorithm.
The paper includes a brief introduction on the history of antiroll tanks followed by the
generation of a mathematical model of a U-tank and application to a sample ship.
Copy weights _
Dither
Plant model
Adaptive
inverse +
controller
+ xk n
+ Ship-tank +
∑w z
i
i
−i
System
Reference +
Sea
waves _
Copy of LMS
Plant model Algorithm εk +
Inverse
modelling
delay
z −∆
Fig. 2 Adaptive inverse model control diagram using filtered-x LMS algorithm
4 SIMULATION AND RESULTS x2 , x& 2 , &x&2 : Sway, sway velocity and sway acceleration
x4 , x& 4 , &x&4 : Roll, roll velocity and roll acceleration
The proposed control strategy is simulated for a x6 , x& 6 , &x&6 : Yaw, yaw velocity and yaw acceleration
particular ship model. The mathematical model of the aij , bij and cij : ship hydrodynamic coefficients
ship tank system used in this study is derived using
Euler method extracted from Lloyd (Lloyd 1989): In this study we have used a simplified version of the
above equations in order to restrict the motion of the
ship to roll only. We therefore have:
Tank
aτ 2 x&&2 + aτ 4 x&&4 + cτ 4 x4 + aτ 6 x&&6 + aτττ&& + bτττ& + cτττ = αP (1) Tank
aτ 4 x&&4 + cτ 4 x 4 + aττ τ&& + bττ τ& + cττ τ = αP (2)
Roll
a 42 &x&2 + b42 x& 2 + ( I 44 + a 44 ) &x&4 + b44 x& 4 + c 44 x 4 + a 46 &x&6 + Roll
b46 x& 6 + c 46 x 6 − [a 4τ τ&& + c 4τ τ ] = F4 ( I 44 + a44 ) &x&4 + b44 x& 4 + c44 x 4 − [a4τ τ&& + c4τ τ ] = F4
where
The simulations were performed for a La Salle class
F4 : Wave moment US Command ship where the specifications of the ship
τ : Tank angle defined as in Fig. 3 are as follows:
P: Pressure of the air provided at the top of the tank
α : Coefficient relating the pressure applied with the
tank dynamics
Table 1 Major parameters of the ship used in the The gains of the inverse controller settle at their final
simulation values (Fig. 5) about 400 seconds after the start of the
simulation (Fig. 6) and the root mean square of the
Maximum Speed 21 knots error (the difference between the actual roll of the ship
Displacement 9600 tonne and the desired roll being zero) gradually reduces (Fig.
Overall Beam 32 m 7).
Length 156 m
Waterline length 150 m wr
Water mass to ship 1.6%
z
The U-tank dimensions used in this simulation are τ
presented in Table 2 where the geometric parameters
are introduced in Fig. 3 and xt is the length of the
hr
tank in the fore/after direction. The tank is situated 1.5 hd
meters above the centre of rotation of the ship. w
Table 2 Dimension of the U-tank Fig. 3 U-tank and its geometric parameters
w 32 m
The gains of the Ship-tank model follow the same
wr 3m pattern and settle at their final values (Fig. 8) about
hr 2m 450 seconds after the start of the simulation (Fig. 9)
hd 1m and the root mean square of the error (the difference
between the actual roll of the ship and roll motion
xt 3.5 m
generated at the output of the ship-tank model given
the same input as the actual system) gradually reduces
(Fig. 10).
The ship-tank system is excited by sea waves generated The roll motion of the ship with the inverse controller
using the wave generating software developed at UCL. and without it is presented in Fig. 4 where considerable
The waves simulate the conditions of sea state 6. roll reduction is observed once the inverse controller
and ship-tank model’s gains are converged and settled
at their final values.
20
10
-5
-10
-15
-20
100 200 300 400 500 600
time(s)
Fig. 4 Roll reduction of the antiroll tank with active U-tank using filtered-x LSM algorithm
0.45 0.045
0.4 0.04
0.35 0.035
0.3 0.03
Amplitude
Amplitude
0.25 0.025
0.2 0.02
0.15 0.015
0.1 0.01
0.05
0.005
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Weight number
Weight number
Fig. 5 Gains of inverse controller after adaptation Fig. 8 Gains of Ship-Tank model after adaptation
(L=15 and ∆=8) (L=15)
0.4 0.06
0.05
0.3 0.04
0.03
Amplitude
Amplitude
0.2
0.02
0.01
0.1
0
-0.01
0
-0.02
-0.1 -0.03
0 100 200 300 400 500 600 100 200 300 400 500 600
time(s) time(s)
Fig. 6 Adaptation of inverse controller’s gains with Fig. 9 Adaptation of Ship-Tank model’s gains with
time (L=15 and ∆=8) time (L=15)
5
Am plitude
5
4 4
3 3
2 2
1 1
0 0
0 100 200 300 400 500 600 700
0 100 200 300 400 500 600 700
time (s)
time (s)
Fig. 7 Root Mean Square error of inverse controller Fig. 10 Root Mean Square error of Ship-Tank
modeling vs. time modeling vs. time
As far as the practical aspects are concerned, the level
of the water inside the tank (denoted by parameter z in 7 REFERENCES
Fig. 3) remains within acceptable limits, see Fig. 11,
similarly for the pump power consumption which never
exceeds 100kW. Abdel Gawad, A. F., Ragab, S. A., Nayfeh, A. H., &
Mook, D. T. 2001, "Roll stabilization by anti-roll
passive tanks", Ocean Engineering, vol. 28, p. 457.
3 Bennett, S. 1991, "Ship stabilization: History," in
2 Concise encyclopedia of traffic and transportation
systems, M. Papageorgiou, ed., Pergamon, New York,
water level (m)
1
pp. 454-459.
0
Broome, D. R. 1998, "Application of ship motion
-1
prediction I", Transactions of Institute of Marine
-2 Engineering, vol. 110, pp. 77-93.
-3
Froude, W. 1861, "On the rolling of ships",
-4 Transactions of Institution of Naval Architects, vol. 2,
0 100 200 300 400 500 600 700
time(s) p. 180.
Fig. 11 level of water inside the U-tanks from Goodrich, G. J. 1969, "Development and design of
equilibrium passive roll stabilisers", Transactions of Royal
Institution of Naval Architects, vol. 111, p. 81.
Lloyd, A. R. J. M. 1989, Seakeeping: Ship behaviour
5 CONCLUSION in rough weather Ellis Horwood Limited, Chichester.
Moaleji, R. & Greig, A. R. 2006, "Development of
It seems that given the conditions of this specific antiroll tanks", Ocean Engineering (in press).
control problem, being the difficulty in modelling the
tank-pump system, the variation of the dynamic of the Parker, M. N. 1965, "Brief review of ship stabilization
ship with weather and loading condition, and the systems", Naval Engineers Journal, vol. 77, no. 4, p.
irregular nature of the sea waves acting as the noise in 640.
this system, filtered-x inverse controller has been able Watts, P. 1883, "On a method of reducing the rolling
to provide a reasonable, affordable and practical of ships at sea", Transactions of Institution of Naval
solution to the problem of roll stabilization of ships at Architects, vol. 24, p. 165.
slow speeds.
Watts, P. 1885, "The use of water chambers for
reducing the rolling of ships at sea", Transactions of
6 AKKNOWLEDGEMENT Institution of Naval Architects, vol. 26, p. 30.
Widrow, B. & Stearns, S. 1985, Adaptive Signal
The financial support provided by the IMarEST Processing Prentice-Hall.
Stanley Gray Fellowship Award is gratefully
acknowledged.