水下机器人抗积分饱和控制及主动容错控制方法

Anti-windup control and active fault tolerant control methods for autonomous underwater vehicles

针对自主式水下机器人(AUV)推力饱和时一般模型控制器控制性能下降问题,提出了基于平滑变参数条件积分法的抗积分饱和控制方法.为适应AUV因抛载等引起的动力学性能变化,提出了一种基于抗积分饱和一般模型控制的主动容错控制方法,同时还研究了传感器故障问题并提出了一种传感器故障容错控制方法.进行了"海狸"号水下机器人试验样机的仿真及水池实验,实验结果验证了所提出的抗积分饱和控制方法的有效性,也验证所提出的容错控制方法在AUV发生故障及动力学性能变化时可有效减小速度、角速度跟踪误差.

To resolve the problem of performance degradation with the generic model controller（GMC） when thrust saturation occurs for an autonomous underwater vehicle（AUV）,an anti-windup control algorithm was devised.It was based on smoothing variations to parameters fed to the conditional integration method.To dynamically adapt the AUV to the impact of changes in loads caused by load rejection,etc.,an active fault tolerant method based on the anti-windup algorithm was developed.Sensor faults were also researched,and a method that is tolerant of fault in the sensor was proposed.Simulations and pool experiments were carried out using the experimental AUV platform named ＂Beaver＂.Experimental results verified the effectiveness of the proposed anti-windup control method.They also showed that the proposed fault tolerant control method can effectively reduce errors of tracking velocity and angular velocity when AUV dynamics change and AUV sensor faults occur.