基于RBF神经网络的AUV路径跟踪分数阶滑模控制  

Fractional-Order Sliding Mode Control Based on RBF Neural Network for AUV Path Tracking

作　　者：王香 张永林[1] WANG Xiang;ZHANG Yong-lin(School of Electronics and Information,Jiangsu University of Science and Technology,Zhenjiang 212003,China)

机构地区：[1]江苏科技大学电子信息学院,江苏镇江212003

出　　处：《水下无人系统学报》2020年第3期284-290,共7页Journal of Unmanned Undersea Systems

摘　　要：针对自主水下航行器(AUV)在对接过程中的路径跟踪问题,提出一种基于径向基函数(RBF)神经网络的分数阶滑模控制算法。首先基于滑模控制设计AUV路径跟踪控制算法,将分数阶微积分引入滑模控制中的等速趋近律以缓解系统的抖振,然后采用RBF神经网络对AUV运动模型中的不确定性及外界干扰进行补偿,最后通过Lyapunov定理证明了控制系统的稳定性。仿真结果表明:所设计的控制器能对AUV路径进行有效跟踪,与传统滑模控制器以及未考虑系统不确定性及外界干扰的分数阶滑模控制器相比,该控制器跟踪速度更快,稳定效果更好,跟踪性能更强。Aiming at the path tracking problem of autonomous undersea vehicle(AUV)in the process of docking,a fractional-order sliding mode control algorithm based on radial basis function(RBF)neural network is proposed.Firstly,the AUV path tracking control algorithm is designed based on sliding mode control,the fractional calculus is introduced into the constant velocity reaching law of sliding mode control to alleviate the shaking of the system.Then,the RBF neural network is used to compensate for uncertainty in the AUV motion model and for external interference.Finally,the stability of the control system is proved by Lyapunov stability theory.Simulation results show that the designed controller can effectively track the path of AUV.Compared with the traditional sliding mode control and the fractional-order sliding mode control without considering system uncertainty and external interference,the proposed control method has faster tracking speed,higher stability and better tracking performance.

关 键 词：自主水下航行器 路径跟踪 径向基函数神经网络 滑模控制 分数阶微积分 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TP242.2[自动化与计算机技术—控制科学与工程]