基于GD-FNN和参考模型的无人艇航向鲁棒自适应控制

Robust adaptive heading control of high-speed USV based on GD-FNN

为使无人艇在复杂环境干扰下能按希望要求改变航向,本文设计了一种基于广义动态模糊神经网络和参考模型的鲁棒自适应控制器。首先针对因环境干扰产生的不确定干扰项基于广义动态模糊神经网络建立了无人艇运动控制的逆动态模型,设计了模糊神经网络的自适应率以进一步调整神经网络权值,并结合无人艇运动控制模型的参考模型设计了无人艇航向鲁棒自适应控制器,然后通过Lyapunov稳定性理论,证明了基于该控制器的无人艇航向控制系统的稳定性,最后采用半物理仿真实验验证了该控制方法的有效性和准确性。

A robust adaptive controller based on generalized dynamic fuzzy neural network is proposed for heading control of a high-speed Unmanned Surface Vessel(USV)under the complex environment.Firstly,the generalized dynamic fuzzy neural network is used to establish the inverse dynamic model of the motion control of USV aimed at the uncertain disturbance generated by environmental interference.Then an adaptive rate of the fuzzy neural network is designed to further adjust the weights of the neural network.Combined with the reference model of the motion control model of the USV,the robust adaptive controller is construct⁃ed.And using the Lyapunov stability theory,the stability of the USV heading control system based on the con⁃troller is proved.Finally,the effectiveness and accuracy of the control method are verified by the semi-physi⁃cal simulation experiment.