考虑输入饱和的基于RBF神经网络补偿的USV滑模控制  

Sliding Mode Control for USV with Input Saturation Based on RBF Network Compensation

作　　者：李敬东 王仁强[1] 刘昌华[1] 陈大伟 LI Jing-dong;WANG Ren-qiang;LIU Chang-hua;CHEN Da-wei(College of Navigation,Jiangsu Maritime Institute,Nanjing Jiangsu 211170,China)

机构地区：[1]江苏海事职业技术学院航海技术学院,江苏南京211170

出　　处：《广州航海学院学报》2018年第3期54-57,共4页Journal of Guangzhou Maritime University

基　　金：江苏省高等学校自然科学研究面上项目(18KJB580003);江苏省高等学校大学生创新创业训练计划项目(201812679021H)

摘　　要：设计了一种考虑输入饱和的USV积分逆推滑模神经网络控制算法.考虑到USV驱动舵机系统输入饱和问题,借鉴抗饱和控制技术设计一种基于内部辅助补偿机制,通过设计RBF网络对补偿机制进行在线逼近.再利用带有积分器的Backstepping方法设计实现USV航向精准控制.同时引入非线性阻尼定律来克服有界外界干扰.最后,采用反推法推导了系统控制律.仿真结果表明,控制器输出平滑,输出对参数摄动不敏感,控制算法有效.A kind of USV course RBF network control algorithm with backstepping was putted forward on the basisof integral backstepping sliding mode. Considering the input saturation problem of the USV drive steering system,ananti-saturation control technology was designed based on the internal auxiliary compensation mechanism,and thecompensation mechanism was applied online by designing the RBF network. Then the Backstepping method withintegrator was used to design and implement USV heading precision control. At the same time,nonlinear dampinglaw was introduced to overcome bounded external interference. Finally,the system control law was derived by usingthe inverse method. The simulation results show that the controller output is smooth,the output is not sensitive toparameter perturbation,and the control algorithm is effective.

关 键 词：USV 跟踪控制 反推法 神经网络 滑模控制 

分 类 号：U666.153[交通运输工程—船舶及航道工程]