AUV path planning, navigation, and control using geophysical data

We propose an integrated motion planning, geophysical navigation, and control (PNC) system for autonomous underwater vehicles (AUVs) that makes explicit use of geophysical information for both planning and navigation. The PNC system exploits the information provided by geomagnetic features observed on the sea-floor to correct the inevitable drift in the vehicle position estimates using dead-reckoning (DR). The three components of the PNC system are: a path planner that enhances the exploratory strength of the Rapidly-exploring Random Tree (RRT) algorithm with transition tests to choose new potential states, a sequential estimation algorithm, and a path following controller. The performance of the system is assessed via hardware in the loop (HIL) simulations to illustrate the significance of using magnetic data for navigation of AUVs. The results obtained are encouraging for inwater tests with an autonomous vehicle of the Medusa class aiming at the validation of the proposed combined system in real-time experiments