Return to Session Menu

DYNAMIC POSITIONING CONFERENCE

October 15-16
15-16, 2013

RISK SESSION

Cooperative Control Applied to Multi-Vessel DP

p
Operations – Numerical and Experimental
p Analysis
y

Asdrubal N. Queiroz Filho and Eduardo A. Tannuri

University of São Paulo
 Cooperative Control Applied to Multi-
Vessel DP Operations - Numerical and
Experimental Analysis
Asdrubal N. Queiroz Filho Eduardo A. Tannuri
University of São Paulo University of São Paulo
São Paulo, SP, Brazil São Paulo, SP, Brazil

October| 2014 1
 Motivation

Bring the ideas of cooperative control

to offshore DP operations that require
several DP Vessels

Subsea Installation
Maintenance Operation

2
Cooperative Control - Motivation

Example – 2 vessels installing a sub-sea equipment

3
Cooperative Control - Motivation

Example – 2 robots manipulating one equipment

4
Cooperative Control - Motivation

Example – 8 robots manipulating one equipment

5
 Cooperative Control - Motivation

Ren et al., Consensus Theory

6
 Objective of the paper

•Present Paper
– Fundamental Analysis using Classic Control Theory;
– Time-domain simulations of a 2-vessel operation using TPN Simulator;
– Experimental results using 1:42 scale models of a DP-PSV;

7
 Mathematical Model

x6A
x2A
x1A

x6B
Linear Dynamics for small yaw angles:
A
x2B
x1B

Y Z
X
B

8
 Non-Cooperative Control
Non-Cooperative
Each vessel with its
𝑑𝑑(𝑥𝑥𝑖𝑖𝐴𝐴 − 𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑖𝑖𝐴𝐴 ) own independent DP
𝐹𝐹𝑥𝑥𝑁𝑁𝑁𝑁 𝐴𝐴
= 𝐾𝐾𝑃𝑃𝑥𝑥 𝑖𝑖 (𝑥𝑥𝑖𝑖𝐴𝐴 − 𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑖𝑖𝐴𝐴 ) + 𝐾𝐾𝐷𝐷𝑥𝑥 𝑖𝑖 + 𝐾𝐾𝐼𝐼𝑥𝑥 𝑖𝑖 �(𝑥𝑥𝑖𝑖𝐴𝐴 − 𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑖𝑖𝐴𝐴 ) 𝑑𝑑𝑑𝑑
𝑖𝑖
𝑑𝑑𝑑𝑑

Vessel A
Set-Point
+ Error
DP A
Thrusts
Vessel A
Horizontal Position
Heading
Calculation -

Measuring
Desired relative
Systems +
position
Filter
between the
vessels

Vessel B
Set-Point
+ Error
DP B
Thrusts
Vessel B
Horizontal Position
Heading
Calculation -

Measuring
Systems +
Filter

9
 Cooperative Control
Cooperative
Measuring
Systems +
Filter
Based on Consensus
-
Control Theory
Vessel A
Set-Point
+ Error
DP A
+ Thrusts
Vessel A
Horizontal Position
Heading
Calculation
-

Kc

Vessel B
+ Error
+ Thrusts Horizontal Position
Set-Point DP B Vessel B
Calculation - + Heading

Measuring
Systems +
Filter

Calculation of
Desired relative relative position
position and and heading
heading error

𝐹𝐹𝑥𝑥𝐴𝐴𝑖𝑖 = 𝐹𝐹𝑥𝑥𝑁𝑁𝑁𝑁
𝑖𝑖
𝐴𝐴
− 𝐾𝐾𝐶𝐶 𝑥𝑥 𝑖𝑖 (𝑥𝑥𝑖𝑖𝐴𝐴 − 𝑥𝑥𝑖𝑖𝐵𝐵 − 𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑖𝑖𝐴𝐴 + 𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑥𝑖𝑖𝐵𝐵 )

10
Cooperative Control

Introduction of
virtual spring and
damper between
the vessels
Vessel B

Vessel A

11
 Case Study - PSV

Property Model Full

Length overall LOA 1.90 m 80.0 m
Full Scale Length between perp. LPP
Beam
1.65 m
0.43 m
69.3 m
18.00 m
Maximum Draft 0.154 m 6.50 m
Maximum Displacement 74 kg 5492 ton

Bow Azimutal
Thruster
P= 883 kW
Main Thruster X=30m Y=0m
P= 6440 kW
X=-40m Y=6m

Main Thruster
P= 6440 kW

Model Scale X=-40m Y=-6m

Stern Tunnel Thruster Bow Tunnel Thruster

P= 883 kW P= 883 kW
X=-28m Y=0m X=30m Y=0m

12
 Numerical Time-Domain
Simulations

October| 2014 13
 Case Study - PSV
Test 1: Tracking Performance
50m 8 m/s wind

B 30° 3.0m 8s
JONSWAP Wave
50m
20°
A
Y 0.5 m/s Current

Objective: To evaluate the ability of

the system to control the relative
distance of the vessels

14
Frequency Domain Analysis
Kc=0
Transfer function dX(s)1/dR(s)1 Kc=0.1*Kp
10
Kc=Kp
0 Kc=10*Kp

Magnitude (dB) -10

-20

-30

-40

-50
45

0
Phase (deg)

-45

-90

-135
-4 -3 -2 -1 0 1
10 10 10 10 10 10
Frequency (rad/sec)

15
 Frequency Domain Analysis
Kc=0
Transfer function dX(s)1/dR(s)1 Kc=0.1*Kp
10
Kc=Kp
0 Kc=10*Kp
Magnitude (dB)

-10

-20
Kc bandwidth
-30

-40

-50

The increase of the gain enlarges the bandwidth of the

system which makes the time response faster.

16
 Frequency Domain Analysis
Kc=0
Transfer function dX(s)1/dR(s)1 Kc=0.1*Kp
10
Kc=Kp
0 Kc=10*Kp
Magnitude (dB)

-10

-20
Kc damping
-30

-40

-50
the total damping of the system is reduced, causing an amplification
in a frequency range near the natural frequency of the system. This
effect is not desirable and imposes an upper limit on the Kc gains

17
 Zero-Pole Map Analysis

0.93 0.88 0.82 0.75 0.68 0.6

System: dXdR System: dXdR
4 Pole : -0.00902 + 0.00498i Zero : -0.00771 + 0.00488i
0.97 Damping: 0.876 Damping: 0.845
Overshoot (%): 0.337 Overshoot (%): 0.698
Frequency (rad/sec): 0.0103 Frequency (rad/sec): 0.00913
2
0.992
Imaginary Axis

0
0.012 0.01 0.008 0.006 0.004

0.992
-2 System: dXdR
System: dXdR
Pole : -0.00902 - 0.00498i Zero : -0.00771 - 0.00488i
0.97 Damping: 0.876 Damping: 0.845
Overshoot (%): 0.337 Overshoot (%): 0.698
-4
Frequency (rad/sec): 0.0103 Frequency (rad/sec): 0.00913

0.93
-6 0.88 0.82 0.75 0.68 0.6

-13 -12 -11 -10 -9 -8 -7 -6 -5 -4

-3
Real Axis x 10

Kc=0.1 x Kp
18
 Zero-Pole Map Analysis
0.06
0.56 0.44 0.32 0.2 0.1
System: dXdR 0.05
Pole : -0.0413 + 0.0598i
0.7 Damping: 0.568 0.04
0.04
Overshoot (%): 11.4
Frequency (rad/sec): 0.0727 0.03
0.84 System: dXdR
0.02 Zero : -0.00132
0.02
Damping: 1
0.95 0.01 Overshoot (%): 0
Imaginary Axis

Frequency (rad/sec): 0.00132

0
System: dXdR
0.95 0.01 Pole : -0.00132
Damping: 1
-0.02
0.02 Overshoot (%): 0
0.84 Frequency (rad/sec): 0.00132
System: dXdR 0.03
Pole : -0.0413 - 0.0598i
-0.04
0.7 Damping: 0.568 0.04
Overshoot (%): 11.4
Frequency (rad/sec): 0.0727 0.05
0.56 0.44 0.32 0.2 0.1
-0.06
-0.04 -0.03 -0.02 -0.01 0 0.01 0.02
Real Axis

Kc=2.0 x Kp
19
 Time Domain Simulation
Relative distances Non cooperative
100 Cooperative
Relative distance set-point
50
X (m)
0

-50
1000 1500 2000 2500 3000 3500 4000

100

50
Y-50 (m)

-50
1000 1500 2000 2500 3000 3500 4000

100
Yaw (deg)

50

-50
1000 1500 2000 2500 3000 3500 4000
time (s)

20
 Time Domain Simulation

Cooperative

Non-Cooperative Vessel A moves backward, in order to

obtain the desired relative distance faster.
The vessel A does not After that, the whole system returns to the
move, and the relative desired position.
motion is only performed
by the vessel B. 21
 Time Domain Simulation

Vessel A effective forces Non cooperative

400 Cooperative
Y Force (kN)

200
0
-200
-400
1000 1500 2000 2500 3000 3500 4000
)

Vessel B effective forces

400
Y Force (kN)

200
0
-200
-400
1000 1500 2000 2500 3000 3500 4000
)

Cooperative

22
 Time Domain Simulation

Surge Motion
Non-cooperative Cooperative

Overshoot 17 % 27 %

Rise-time 56 s 15 s
10% settling time 248 s 50 s

23
 Case Study - PSV

Test 2: Disturbance rejection 4.0m 12s

JONSWAP

B
15°
0.5 m/s Current
50m 30°

Y
X

Objective: To evaluate the ability to

keep the relative distance, despite of
the environmental conditions

24
 Frequency Domain Analysis
50

Magnitude (dB)
-50

-100
Slow drift range
-150
0
Increase of Kc attenuates the Kc=0

system response near the slow Kc=0.1*Kp

Phase (deg)

-90 Kc=Kp
drift frequency range which is Kc=10*Kp
a desirable effect since it is -180
wanted that the system rejects
the slow drift movement. -270
-5 -4 -3 -2 -1 0 1
10 10 10 10 10 10 10
Frequency (rad/sec)

25
 Zero-Pole Map Analysis
-3
x 10 Pole-Zero Map
8
0.993 0.986 0.972 0.945 0.88 0.65

System: dXdD
6
0.998 Pole : -0.007 + 0.00714i
Damping: 0.7
4 Overshoot (%): 4.6
System: dXdD Frequency (rad/sec): 0.01
0.999 Pole : -0.07
2 Damping: 1
Overshoot (%): 0
Imaginary Axis

Frequency (rad/sec): 0.07

8 0.07 0.06 0.05 0.04 0.03 0.02 0.01
0

-2
0.999
System: dXdD
-4 Pole : -0.007 - 0.00714i
Damping: 0.7
0.998 Overshoot (%): 4.6
-6
Frequency (rad/sec): 0.01

0.993 0.986 0.972 0.945 0.88 0.65

-8
-0.08 -0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0
Real Axis

Kc=0.1 x Kp
26
 Zero-Pole Map Analysis
Transfer function dX(s)1/dD(s)1
0.06
0.54 0.4 0.3 0.21 0.13 0.06
System: dXdD 0.05
Kc=2*Kp
Pole : -0.0413 + 0.0598i
0.7
Damping: 0.568 0.04
0.04
Overshoot (%): 11.4
Frequency (rad/sec): 0.0727 0.03
System: dXdD
Pole : -0.00132 0.02 Dominant
0.02 0.9
Damping: 1
0.01
(slow pole)
Overshoot (%): 0
Imaginary Axis

Frequency (rad/sec): 0.00132

0

0.01
-0.02 0.9
0.02

System: dXdD 0.03

Pole : -0.0413 - 0.0598i
-0.04
Damping: 0.568 0.04
0.7
Overshoot (%): 11.4
Frequency (rad/sec): 0.0727 0.05
Kc=2 x Kp -0.06
-0.045 -0.04
0.54

-0.035 -0.03
0.4

-0.025 -0.02
0.3 0.21

-0.015
0.13

-0.01
0.06

-0.005 0
Real Axis

27
 Time Domain Simulations
4.0m 12s

Improvement in the JONSWAP

stationkeeping
B
15°
0.5 m/s Current
50m 30°

performance
A

Y
X

Non cooperative
Cooperative
90 3 90 2.5
120 60 90 4
120 60
2 120 60
2
3
1.5
150 30 150 30
1 1 150 2 30

0.5 1

180 0
180 0
180 0

210 330
210 330 210 330

240 300
240 300 240 300
270
270 270

Maximum relative position error (m) 28

 Time Domain Simulation

Non-Cooperative Cooperative

The vessel A motion does Vessel A moves backward, in order to

not affect the vessel B obtain the desired relative distance faster.
motion and vice-versa.
After that, the whole system returns to the
desired position. 29
 Experimental Results

October| 2014 30
 Experimental Set-up

University of São Paulo – Academic Tank

wind

waves X
Z
20m x 8m x 2m
Y

31
 Experimental Set-up

Experimental Set-up for Tests with 2 DP Vessels

Nypos Cameras

Model 1 Model 2

Thruster Commands
NYPOS Computer

X,Y,ZZ (2 vessels)

Control Computer

32
 Experimental Set-up

Driver DC motor
Radio Computer

Small scale
Modelo DP Vessel
Rebocador
Driver step motor
Driver step
motor
Driver DC motor

Radio
DC Supply
RS-485 Comm.
Controller

Driver DC motor

Driver step motor

Embedded
Electronics

33
 Experimental Set-up

3 cameras

Position In-House DLL for real-time communication

Measurement Gigabit Network

34
 Experimental Set-up
Control
Software

Option 2: Matlab/Simulink based

control, adequate for research
project.
Option 1: DP-BR, a complete software for 1-
vessel control, using PID control, Kalman Filter
and Lagrangean Based Thruster Allocation

35
 Experimental Results
0

Test 0 – Non-cooperative control -0.1

X Position(m)
-0.2 Vessel A

same PID controller gains for both vessels -0.3

Vessel B
Set-Point

-0.4

significant differences between the

140 160 180 200 220 240 260 280 300

performances of the vessels, mainly in sway -0.5

and yaw directions -0.6

Y Position(m)
-0.7 Vessel A
Vessel B (+0.9m)

It can be explained by some discrepancies in -0.8 Set-Point

the ballasting arrangements and the -0.9

200 220 240 260 280 300 320

propellers mechanical construction and the

final rpm-force relation.
40
Vessel A
30

Yaw Angle (deg)

Vessel B
20 Set-Point

A 10

B -10
0.2m

300 320 340 360 380 400 420 440

Y 10o and Time(s)
0.2m 30o
X

36
 Experimental Results
Test 0 – Non-cooperative control

380 400 420 440

me(s) Non-Cooperative

Even small differences between vessels

induce large relative motions in DP vessels
37
 Cooperative Control
Test 1 – Cooperative control – Yaw maneuver
K Cx6 = 0 K Cx6 = 5 K Cx6 = 8 K Cx6 = 12
40 Vessel A
Vessel B
Y
30 Set-Point
X Vessel A

Yaw Angle (deg)

20
0.9m 30º
10
Vessel B
0

-10
40 60 80 100 120 140 160 180 200 220
The higher is the gain the closer Time (s)

are the two curves during the yaw 40

K Cx6 = 0 K Cx6 = 15

maneuver 30
Vessel A
Vessel B
Set-Point
Yaw Angle (deg)

Other paper of the authors (to be

20

presented at IFAC-CAMS 2013)

10

demonstrates the stability analysis

0

and gain sensitivity

-10
30 40 50 60 70 80 90 100 110 120
Time (s)

38
 Cooperative Control
Test 1 – Cooperative control – Yaw maneuver

=8 K Cx6 = 12
Vessel A
Vessel B
Set-Point

160 180 200 220

Time (s)

Cooperative
The faster vessel slow down to follow the slower vessel!
Small relative motion 39
 Cooperative Control
Test 1 – Cooperative control – Yaw maneuver

K Cx6 = 0 K Cx6 = 5 K Cx6 = 8 K Cx6 = 12

1.1
Y
1
X Vessel A

Relative Distance (m)

0.9
0.9m 30º
0.8

Vessel B Midship
0.7 Bow
Stern

40 60 80 100 120 140 160 180 200 220

Time(s)

Distance between points

40
 Cooperative Control
Test 2 – Cooperative control – Sway/Yaw maneuver

Vessel A
Y Vessel B
-0.6 Set-Point
X Wind and Wave

Y position (m)
direction -0.7
Vessel A
-0.8
30º -0.9
0.2m 0.9m
-1
50 100 150 200 250 300 350

Kc2=0
Time (s)
Vessel B
0.14

0.12
Kc2=12 Kc2=8 Kc2=10
0.07
Absolute relative distance error (m)
Absolute relative distance mean error

0.1
0.06
0.08
0.05

0.04 0.06
(m)

0.03
0.04

0.02
0.02
0.01
0
0.00 50 100 150 200 250 300 350
0 2 4 6 8 10 12 14 Time (s)

Kcx2

41
 Cooperative Control
Test 2 – Cooperative control – Sway/Yaw maneuver
With Environmental Action

-0.2 Vessel A

Y position (m)
Vessel B
-0.4
Set-Point
-0.6
Cooperative Non-cooperative Midship
Bow -0.8
Stern
1.1
-1
150 200 250 300 350 400 450 500 550
Relative distance (m)

1
Time (s)
Cooperative Non-cooperative Cooperative
0.9

40
Yaw Angle (deg)

0.8

0.7
20
0.6
150 200 250 300 350
Time (s)
0

150 200 250 300 350 400 450 500 550

Time (s)

42
 Numerical experimental comparison
Non-cooperative Cooperative

Experimental

Time (s)

Non cooperative control Cooperative control

50 50
Vessel A Vessel A
Yaw (deg)

Yaw (deg)
Vessel B Vessel B
Numerical 0 0

-50 -50
3000 3200 3400 3600 3800 4000 2800 3000 3200 3400 3600 3800 4000
time (s) time (s)

43
 Cooperative Control

Cooperative control based on the consensus concept applied

to coordinated manoeuvres.

Numerical tests showed that the tracking performance and

disturbance rejection of the relative positioning can be
adjusted by a proper selection of cooperative control gain.

The next steps ....

•To derive the model of the n-vessel system

•To study the effects of communication problems

44
 Cooperative Control

The next step 1) To derive the model of the n-vessel system

and to test in the larger wave basin of TPN

45
 Cooperative Control

The next step 1) To study the effects of communication

problems using the TPN-USP full bridge simulator

46
Thank you

Asdrubal N. Queiroz Filho

[email protected]

Eduardo Aoun Tannuri

[email protected]

47