基于自适应区域跟踪的自主式水下机器人容错控制

研究自主式水下机器人(autonomous underwater vehicle, AUV)的推进器自适应区域跟踪容错控制方法。 与传统的自主式水下机器人容错控制方法不同,采用区域跟踪控制思想,将控制目标设定为以期望轨迹为中心的空间区域。 针对系统中存在的不确定性及推进器故障问题,采用神经网络进行在线辨识。 考虑到推进器故障时存在推力饱和而导致神经网络学习发散的问题,提出一种包含饱和因子的神经网络权值调整方法。 通过仿真,对所提方法的有效性进行验证。

Fault-tolerant control of autonomous underwater vehicle based on adaptive region tracking

An adaptive region tracking fault-tolerant control for the thrusters of autonomous underwater vehicle was proposed. Different from the traditional fault-tolerant control methods of autonomous underwater vehicle, the region tracking control theory was adopted, and the control target was designed as a spatial region. For the uncertainty and thruster fault in the system, the neural network was used to identify them online. Considering the problem of the divergence of neural network caused by the thrust saturation during the thruster fault, a neural network weight adjustment method based on a saturation factor was proposed. The effectiveness of the proposed method was verified by simulation.