Design of a Vessel Dynamic Location System Observer Using the Neural Network

This research develops a neural network observer for a vessel dynamic positioning system with unknown dynamics and disturbances. The location readings for vessel dynamic positioning systems are often accessible, but the velocity values are not. Additionally, sounds often tamper with the measured location and heading data, which might hurt the effectiveness of the dynamic location system. With the use of the corrupted location data, the developed observer is shown to be able to estimate full of the dynamic system's state, which including location and velocity data. It is shown that the observer estimating errors for the DP system's state are uniformly finally limited using the Lyapunov theory. Additionally, the results show that the constructed neural network observer can provide accurate velocity and location estimations from distorted location data. The efficacy and performance of the suggested neural network observer is shown using an example.