Autonomous Underwater Pipeline Tracking Control Based on Visual Images

Abstract

In order to realize the underwater pipeline tracking with an Autonomous underwater vehicle (AUV) under environment disturbances, a hybrid control method is presented based on vision navigation information of the monocular camera. Since the original image may be distorted, the information of the pipeline in images needs to be transformed first. Moreover, a robust controller plays an important role in the motion control of an AUV in disturbances. Therefore, a DRNN (Diagonal Recurrent Neural Networks) and S nonlinear hybrid controller is established. Compared with the PD controller, the result shows that the controller has smaller steady-state error and better self-adaptive capacity. The simulation results show that the control method is effective with strong robust and adaptive capability for the process for autonomous underwater pipeline tracking with fixed tracking height from the pipeline. Moreover, the method removing the distortion of images is concise and efficient, and the parameters of the controller are easy to be adjusted.