Design of integrated navigation system for autonomous underwater robotic

Underwater robotic navigation system is one of challenge issues of autonomous underwater robotics, and it is also the underwater robotics´ difficult field and active research field. An inertial integrated navigation system, which is based upon inertial measurement unit, Doppler velocity log, TCM2, GPS and ultra-short base line, is designed for underwater robotics in this paper. The configuration and measurement sensors of the navigation system are analyzed, and an estimation algorithm based on multi-layer Kalman filtering is presented in particular. Furthermore, in order to deal with the position information delay of sensors, dead-reckoning estimation algorithm and double-layer estimation algorithm are proposed. Simulation results validate the correctness and usability of the integrated navigation system and the validity of these algorithms.