Research on a Terrain-Aided Navigation Method Based on a Particle Filter and Its Improved Algorithm

As traditional navigation methods cannot meet the requirements of the long endurance navigation of autonomous underwater vehicles, this paper used particle filter algorithm for terrain-aided navigation method. Simulation experiments show that the particle filtering algorithm has higher accuracy compared with traditional SINS/DVL integrated navigation methods. When the nonlinear measurement model was applied, especially at the inflexion point where the sign of the slope changed, the uncertainty of the measurement was the key reason for the degradation and divergence of the filter. To reduce the ambiguity of the local measurement in particle filtering, an improved particle filtering algorithm was developed and analyzed by simulation. The simulation results verified the effectiveness of this method for navigation of a seabed terrain.