Small Autonomous Underwater Vehicle Navigation System Based onAdaptive UKF Algorithm
基于自适应UKF算法的小型水下机器人导航系统

Ocean current disturbance and attitude, heading errors can cause uncertain navigation system model error. To solve the above problem, an adaptive unscented Kalman filter (AUKF) with model error is designed for a small autonomous underwater vehicle's (SAUV) dead reckoning (DR) navigation system. Firstly, three-dimensional motion of SAUV continuous time model is designed. Then, the proposed AUKF algorithm is deduced according to maximum a posterior (MAP) theory. Finally, simulation results show that the algorithm can overcome the model error caused by disturbance currents and attitude, and heading errors. Compared with the conventional UKF algorithm, the filter precision of the SAUV's DR navigation system in complex sea state is improved a lot by adopting the proposed algorithm.