基于L2干扰抑制的水下机器人三维航迹跟踪控制
Three-dimensional path tracking control for an autonomous underwater vehicle based on L-two disturbance attenuation method
摘要点击 1895  全文点击 1915  投稿时间:2010-03-10  修订日期:2010-06-09
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/j.issn.1000-8152.2011.5.CCTA100229
  2011,28(5):645-651
中文关键词  自治水下机器人  三维航迹跟踪  神经网络控制  L2干扰抑制
英文关键词  autonomous underwater vehicle  three-dimensional path tracking  neural network control  L-two disturbance attenuation
基金项目  国家自然科学基金资助项目(60704004).
作者单位
张利军 哈尔滨工程大学 自动化学院 
贾鹤鸣 哈尔滨工程大学 自动化学院 
边信黔 哈尔滨工程大学 自动化学院 
严浙平 哈尔滨工程大学 自动化学院 
程相勤 哈尔滨工程大学 自动化学院 
中文摘要
      为实现自治水下机器人(AUV)的三维航迹跟踪控制, 考虑了非线性水动力阻尼对AUV系统的影响和外界海流干扰作用, 提出了基于L2干扰抑制的鲁棒神经网络控制方法. 该方法基于李雅普诺夫稳定性理论, 设计神经网络控制器补偿非线性水动力阻尼和外界的海流干扰, 再将神经网络的估计误差当做AUV系统的外部干扰用L2干扰抑制控制器予以消除. 最后针对某AUV进行了螺旋线三维下潜跟踪控制仿真实验, 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AUV三维航迹的精确跟踪.
英文摘要
      In the three-dimensional path tracking of an autonomous underwater vehicle (AUV), we consider the external current disturbances and the nonlinear hydrodynamic damping effects of the AUV, and propose a robust neural network control based on Lyapunov stability theory for L-two disturbance attenuation. A neural network controller is designed to compensate for the nonlinear hydrodynamic damping and external currents disturbances,and the estimated error of neural network is eliminated as external disturbances in AUV system by the L-two disturbance attenuation controller. A threedimensional path tracking simulation is carried out for an experimental AUV. The simulation results show that the designed controller suppresses the influence of time-varying nonlinear hydrodynamic damping to the AUV system, and attenuates the external currents disturbances as well.