考虑执行器故障的无人帆船事件触发控制

针对实际海洋环境下, 无人帆船在艏向跟踪控制任务中存在的模型参数未知、控制输入频繁抖振和执行器磨损等问题, 提出一种考虑执行器故障的无人帆船事件触发控制算法。首先, 采用径向基函数神经网络(radius based function neural networks, RBF-NNs)对系统的未知模型参数进行在线逼近。其次, 在无人帆船艏向数学模型中引入执行器故障模型, 并且在艏向控制器设计中考虑帆结构造成的转船力矩, 设计基于事件触发机制的艏向控制律来减少控制输入的频繁抖振和执行器磨损。最后, 通过李雅普诺夫稳定性判据证明了所提控制算法满足半全局一致有界(semi-global uniform ultimate bounded, SGUUB)稳定。数值仿真结果表明, 相比于传统的艏向控制算法, 所提算法能够在保证艏向控制性能的基础上, 极大地降低控制输入的频繁抖振和减少执行器的磨损。

Event-triggered control for unmanned sailboat with actuator failures

Aiming at the heading control problem for the unmanned sailboat with the unknown model parameters, the chattering frequently of control input and the actuator wear under the practical marine environment, an event-triggered control algorithm with the actuator failures is proposed. Firstly, the radius based function neural networks (RBF-NNs) is introduced to approximated online the unknown model parameters of the system. Secondly, the actuator failure model is introduced into the heading mathematical model of unmanned sailboat, and the turning torque caused by sail structure is considered in the design of heading controller. A heading control law based on event triggering mechanism is designed to reduce the frequent chattering of control input and actuator wear. Finally, through the Lyapunov stability criterion, it is proved that the proposed control algorithm satisfies semi-global uniform ultimate bounded (SGUUB) stability. Numerical simulation results show that compared with the traditional heading control algorithm, the proposed algorithm can greatly reduce the frequent chattering of control input and the wear of actuator on the basis of ensuring the heading control performance.