SINS/GNSS-Integrated Navigation of Surface Vessels Based on Various Nonlinear Kalman Filters and Large Ship Dynamics

Abstract

Generally, studies investigating the strapdown inertial navigation system/global navigation satellite system (SINS/GNSS)-integrated navigation of surface vessels focus on adaptive strong nonlinear algorithms to handle the system and measurement noise induced by the complex environment and motion state. However, these studies rarely consider the suitability of strong nonlinear adaptive optimization in all conditions or the existence of any restriction. The application conditions for these standard nonlinear filters with respect to the surface vessel SINS/GNSS require investigation. In this study, the estimation accuracy of the motion state obtained using the extended Kalman filter, unscented Kalman filter, and cubature Kalman filter in case of different vessel motions is compared based on the simulated ship sensor data obtained using a dynamic large surface vessel model under various marine conditions. Compared with previous studies, the data generator in this study simulates the actual ship movements under various conditions, based on which considerably detailed and practical analysis and conclusion can be realized in case of the SINS/GNSS-integrated navigation obtained using various standard nonlinear filters. In particular, the situations often encountered by large surface vessels during integrated navigation attributed to environmental interference or instrument failure, including system noise amplification, initial alignment error, and GPS outage, are investigated.