自治式水下机器人推进器故障检测、分离与重构

研究自治式水下机器人推进器故障检测、分离与重构问题。针对水下机器人故障诊断残差法中残差阈值不易选取的问题,提出了一种基于观测器的水下机器人推进器故障检测与分离方法,通过构建故障检测观测器对推进器故障与残差信号进行解耦,使推进器出现故障后仅引起与其相关的残差呈单调变化,从而可选择较大的阈值进行故障检测以提高诊断系统的可靠性,该方法在故障检测的同时可进行故障分离。针对系统不确定性导致的推进器故障重构精度低的问题,提出了一种RBF神经网络与等效输出注入相结合的故障重构方法,采用等效输出注入方法估计不确定性项,从神经网络输出中重构推进器故障。

Thruster Fault Detection,Isolation and Reconstruction for Autonomous Underwater Vehicle

Thruster fault detection,isolation and reconstruction for autonomous underwater vehicle is studied.As residual threshold is difficult to select in residual method of underwater vehicle fault diagnosis,a fault detection and isolation method of underwater vehicle thrusters based on observer is proposed.Thruster fault is decoupled from residual signal through constructing fault detection observer. Then only the related residuals appear monotonous changes after thruster failure.Thus,a larger threshold is select...