基于BLF的半潜式海洋平台自适应神经网络容错控制

Adaptive Neural Network Fault-tolerant Control for Semi-submersible Ocean Platforms Based on BLF

针对具有不确定性和故障半潜式海洋平台,提出了一种反步自适应神经网络容错控制方法。基于半潜式海洋平台低速状态下的动力定位模型,利用径向基函数(radial basis function,RBF)神经网络对系统的不确定性集合进行在线逼近,从而对控制器进行补偿;通过设计基于障碍李雅普诺夫函数(barrier lyapunov function,BLF)的自适应律,实现对系统的状态约束;结合反步法设计了一种鲁棒定位与跟踪控制方法。通过稳定性理论证明了该闭环系统是稳定的,仿真结果表明,该控制器能使半潜式海洋平台完成动力定位和故障情况下的轨迹跟踪任务。

Aiming at the semi-submersible ocean platforms with uncertainties and faults,a backstep adaptive neural network fault-tolerant control method is presented.Based on the dynamic positioning model of semi-submersible ocean platforms at low speed,the radial basis function(RBF)neural network is used to approximate the system uncertainty set online,and to compensate the controller.Secondly,an adaptive law based on Barrier Lyapunov Function(BLF)is designed to realize the constraint of the state of the system.Finally,a robust positioning and tracking control method is designed combining with back-step method.The closed-loop system is proved stable by the stability theory,and the simulation results show that the controller can accomplish dynamic positioning and trajectory tracking under fault conditions of semi-submersible ocean platforms.