Autonomous Underwater Vehicle Navigation Using an Adaptive Kalman Filter for Sensor Fusion

In order to improve the navigation accuracy of an autonomous underwater vehicle, a terrain passive integrated navigation system, which consists of strapdown inertial navigation system, underwater terrain map, bathometer, magnetic compass and doppler log is presented. In the integrated system, the position measure equation of terrain match algorithm is as measurement equation, which adding to the doppler velocity measure equation and magnetic compass course measure equation. And the parameters are estimated using UD factorization-based adaptive kalman filter, which adjust the measure noise to suit change of measure course, inhibiting the error divergence of navigation system. The simulation results prove that the terrain passive integrated navigation system, as number of navigation device is constant, can reduce the attitude angle error and location error of underwater vehicle effectively, and demand the requirement of high precision and low cost.