An Effective Motion Control Based on 2-DOF PID and ELM for AUV

In the complex sea conditions, autonomous underwater vehicle (AUV) need to perform some high-precision tasks, which requires the motion control of AUV have the best command tracking performance, anti-interference ability and robustness. We adopt an effective feed-forward two degree-of-freedom (2-DOF) PID control to ensure the motion control system of AUV have better anti-interference ability and command tracking performance, and use Extreme Learning Machine (ELM) to learn and train corresponding parameters set in order to obtain appropriate parameters quickly and reduces the cumbersome process of artificially adjusting the parameters. The horizontal experiments have been carried out in December 2017 and April 2018 at TuanDao bay, QingDao using own AUV platform, Sailfish.