基于有限时间干扰观测器的改进模型水下机器人自适应鲁棒容错控制

针对无人水下机器人（unmanned underwater vehicle ，UUV）工作中存在的执行器故障，在系统不确定性与外界干扰下，提出一种基于有限时间扰动观测器（finite time disturbance observer ，FTDO），并结合改进模型的自适应鲁棒容错控制方法。一方面，FTDO能在有限时间内对外界环境干扰进行估计；另一方面利用滑模控制加上径向基神经网络（Radial Basis Function Neyral Network，RBF） 的万能逼近特性，建立带有执行器故障的输入补偿；其中改进模型的引入解决了系统不确定性导致的输入饱和，提高了稳定性与鲁棒性；其次采用一种新型的双幂趋近律使滑模量在更短时间收敛到稳态误差界内；仿真与水池实验结果表明了所提方法相对于滑模控制有着更好的容错效果。

Improved model unmanned underwater vehicle adaptive robust fault-tolerant control based on finite time disturbance observer

Aiming at the actuator failure in the work of unmanned underwater vehicle (UUV), an adaptive robust fault-tolerant control method based on finite time disturbance observer (FTDO) and improved model is proposed under system uncertainty and external interference. On the one hand, the external environmental interference can be estimated by FTDO in a limited time; On the other hand, sliding mode control and the universal approximation characteristics of Radial Basis Function Neyral Network(RBF) are used to establish input compensation with actuator fault. Among them, the input saturation caused by system uncertainty is solved by the improved model, which improves the stability and robustness. Secondly, a new type of double-power approximation law is adopted to make the sliding modulus converge to the steady-state error range in a shorter time. The simulation and pool experimental results show that the proposed method has a better fault tolerance effect than the sliding mode control.