Bathymetry and Atomic Gravimetry Sensor Fusion for Autonomous Underwater Vehicle

Terrain-aided navigation provides a drift-free navigation approach for autonomous underwater vehicles. However, velocity is often tricky to estimate with conventional bathymetry (mono or multi-beam telemetry) sensors. Cold atom gravimetry is a promising absolute and autonomous additional sensor that is seldom considered for this kind of application. We investigate a multi-beam telemeter and gravimeter centralized fusion scenario and the resulting observability gain on velocity. To do so, an Adaptive Approximate Bayesian Computation Regularized Particle Filter is implemented and compared to conventional Regularized Particle Filter. Numerical results are presented and the robustness of the bathymetry and gravimetry fusion strategy is demonstrated, yielding less non-convergence cases and more accurate position and velocity estimation.