||GPS technology has been developed quite mature and can quickly provide accurate positioning of most Earth surface. GPS has wide range of applications, from daily life to military purposes. But it still has limitations. Because GPS signals can only pass through shallow waters, underwater positioning is usually achieved by using the transponders. The transponders, which are laid on the seabed or underwater vehicle, receive the signals transmitted by the transceiver and responses back to transceiver. Knowing the time difference between the transceiver transmitting signals and receiving signals, the position can be calculated with speed of sound in water. Nevertheless, the sea is not a homogeneous flow field and the speed of sound at different water densities varies. As a result, the calculated position is not accurate enough. Therefore, this paper aims at location prediction for the underwater towed vehicle using computer simulation technique.|
The underwater towed vehicle links a surface ship with cable or rope. The towed vehicle itself does not have propulsive power. The boat by moving drags the rope and changes the underwater towed vehicle’s position by pulling. In order to estimate the position and motion of the underwater vehicle, it is required to fully understand the motion of the cable. Based on the mechanics and motion analysis of the cable, the configuration and motion of the cable can be obtained. This information will be applied to predict future movement of the cable and the underwater towed vehicle. In this thesis, three types of actual data paths, including straight lie, L-type, and U-type are investigated. Simulation results are compared with real data. The simulation speed can be completed in 12 seconds for about 200 cable configurations. Length of the underwater cable is from several hundreds to nearly a thousand meter, and average distance error between the predicted position and the corresponding actual position is about 50 meters. The presented simulation technique successfully demonstrates acceptable prediction performance with fast computational efficiency.