基于神经网络的水下机器人三维航迹跟踪控制

本文研究了水下机器人三维航迹跟踪控制问题.在充分考虑了模型中不确定水动力系数和外界海流干扰的基础上,提出了基于神经网络的自适应输出反馈控制方法.控制器由3部分组成:基于动态补偿器的输出反馈控制项、神经网络自适应控制项和鲁棒控制项.神经网络所需的自适应学习信号由线性观测器提供.基于Lyapunov稳定性理论证明了控制系统的稳定性.最后针对某AUV进行了空间三维航迹跟踪控制仿真实验,结果表明设计的控制器可以较好地克服时变非线性水动力阻尼对系统的影响,并对外界海流干扰有较好的抑制作用,可以实现三维航迹的精确跟踪.

Three-dimensional path tracking control for autonomous underwater vehicle based on neural network

Considering the uncertain hydrodynamic coefficients and environmental current disturbances, we propose an adaptive output feedback neural network controller for the three-dimensional path tracking control of an autonomous underwater vehicle (AUV). The controller consists of three control terms: the output feedback control term of the dynamic compensator, the neural network adaptive control term and the robust control term. An observer is designed to provide the neural network with the adaptive learning signal. The stability of the controller is analyzed by Lyapunov theory. A threedimensional path tracking simulation is carried out on an experimental AUV; simulations results show that the designed controller provides high tracking precision, effectively suppresses the effect of the time-varying nonlinear hydrodynamic damping on the AUV system, and significantly attenuate the external currents disturbances.