|
4. VERIFICATION IN MODEL TESTS FOR CONTROL PERFORMANCE VERIFICATION BY SIMULATION
4.1 Facilities and tested models
Experiments were carried out to confirm the applicability of dynamic positioning using neural networks at the seakeeping and maneuvering basin of Mitsubishi Heavy Industries' Nagasaki R&D Center.
(1) Tested models
The scale ratio of the models was set at 1/50, considering the capability of the basin equipment. The principal dimensions for the models and the corresponding actual vessels are shown in Table 2.
(a) FPSO:
Type: 120 kdwt tanker
Actuator: 30 ton rotable type thruster
(b) Shuffle tanker
Type: 68 kdwt tanker
Actuator: None
(c) Hawser
Length:50m
(2) Test arrangements
Waves and current were considered as external forces in the experiments. The waves were generated by a multi-directional wave maker and current conditions were created by moving the towing carriage.
4.2 Measurement items
The measurement items consist of the motions of the FPSO and the shuttle tanker models, the thrust of the rotable thruster installed on the FPSO model, and the wave. The positions of the FPSO and the shuttle tanker models were obtained from the towing carriage position and the relative distance between the models and the towing carriage by means of optical position measurement equipment. Moreover, talet mooring force acting on the FPSO model was obtained from load cells attached to the experimental mooring equipment.
4.3 Experiment conditions
The control and the external force conditions for the experiments are shown in Table 3. The conditions of significant wave height were 3.5m and 4.5m by adopting the Jonswap Spectrum; current directions were -15 degrees, 5 degrees, 15 degrees and 25 degrees on the basis of the wave direction; and conditions of current velocity were 2 knots and 3 knots. The experiments were carried out in the condition of non-control, PID control and neural network control in which the control target was the rotating angle of the FPSO model thruster.
Table 2 Principal dimensions of FPSO and shuttle tanker models
I/S=1/50,  =7.071
| |
FPSO |
Shuttle tankaer |
| Load condition |
83% Load |
30% Load |
| |
Prototype |
Model |
Protoype |
Model |
| Length L |
255.7 m |
5,114 mm |
230.0 m |
4,600 mm |
| Breadth B |
42.31m |
846.1mm |
37.51 m |
750.9 mm |
| Depth D |
22.00 m |
440.0 mm |
21.25 m |
425.0 mm |
| Draught d |
13.30 m |
266.0 mm |
8.60 m |
172.0 mm |
| Displacement Δa |
121,886 t |
951.31 kg |
61,004 t |
476.13 kg |
| Trim t |
0.0% |
0.22% Aft |
| Height of center of gravity KG |
13.00 m |
260.0 mm |
10.68 m |
213.6 mm |
| Metacentric height GM |
6.00 m |
120.0 mm |
7.03 m |
140.6 mm |
Hawser connected position XH
(from midship) |
148.5 m |
2,970.0 mm |
115.0 m |
2,300.0 mm |
Mooring point x0-F,y0-F
(from midship) |
38.4m, 0.0 m |
767.1 mm, 0.0 mm |
|
|
|
4.4 Results of the experiment
Figs. 9-11 show loci comparison of the FPSO and the shuttle tanker, time histories of hawser tension, and the maximum and average values of hawser tension between 3.5m and 4.5m. The top and the bottom of each figure show the results for 3.5m and 4.5m significant wave height conditions respectively. The figures also include the comparison of no control, PID control and neural control.
In terms of comparison of loci, there is no notable difference according to control method or wave height, since the quantity is comparatively small, as shown in Fig.9.
Table 3 External force condition
| |
Prototype |
Model |
| Wave |
Significant wave height H1/3 |
4.5m
3.5m |
9.0cm
7.0 cm |
| Peak period TP |
6.5s |
0.92s |
| Spectrum |
JONSWAP |
| Directional distribution function |
COS4β |
| Wave direction βW |
0 deg |
| Current |
Current velocity UC |
2kn
3kn |
O.146 m/s
0.218 m/s |
| Current direction ψC |
-15,5,15,25deg |
|
Fig.9 Comparison of loci
Significant wave height: 3.5m
Significant wave height: 4.5m
|
