Analysis of data fusion algorithms for the vessel with the dynamic positioning system

The dynamic positioning (DP) system on the vessel is operated to control the position and heading of the vessel with the use of propellers and thrusters installed on the board. On DP vessels redundant measurement systems of position, heading and the magnitude and direction of environmental forces are required for safety at sea. In this case, a fusion of data is needed from individual measurement devices. The article proposes a new solution data fusion algorithm of particle Kalman filter as a cascade combination of particle filter and extended Kalman filter. The estimation quality of the proposed data fusion algorithm is analysed in comparison with the classic: extended Kalman filter (EKF), nonlinear observer (NO), and particle Kalman filter (PKF). Simulation studies were executed for emergency scenarios to evaluate the robustness of the algorithm analyses to measurement errors.