近水面AUV自适应输出反馈控制器设计  

Adaptive Output Feedback Controller Design for AUV in Shallow Water Area

作　　者：张利军[1] 齐雪[1] 赵杰梅 庞永杰[1] 

机构地区：[1]哈尔滨工程大学自动化学院,哈尔滨150001

出　　处：《中国造船》2012年第2期51-61,共11页Shipbuilding of China

基　　金：国家自然科学基金(60704004);哈尔滨工程大学水下机器人技术国防科技重点实验室基金;黑龙江省青年骨干教师计划;中国教育部博士基金(20102304110003)

摘　　要：近水面航行的自治水下机器人(AUV)受到较强的波浪干扰,此干扰可进入测量输出信号,影响AUV的实际位置和姿态。若此干扰进入控制设计环节,必然增大控制输出,降低控制精度、浪费燃料,严重时可破坏驱动装置。针对此问题,设计了自适应输出反馈控制器,用以实现强干扰下的AUV水平面轨迹跟踪控制。首先设计观测器将系统输出信号中的干扰过滤掉,使其无法进入控制设计环节。然后采用局部逼近的径向基神经网络对不确定水动力系数和径向、横向及艏向角速度耦合产生的非线性结构进行补偿。利用Lyapunov稳定性理论证明闭环系统的稳定性。仿真实验表明控制器对粘性阻尼系数和未知非线性结构具有很好的自适应性;对附加质量不确定性和干扰的变化有很强的鲁棒性。The autonomous underwater vehicle （AUV） will be disturbed by wave motion when operating in shallow water area. The significant disturbance due to wave motion will be introduced into the measurable output signals and will influence the actual position and attitude of AUV. If the wave disturbance enters into the control loop, it will cause the increase of the output control, reduction of the control accuracy, waste of fuel and wear of actuators in serious situation. According to this problem, an adaptive output feedback controller is designed to realize the plane trajectory tracking control of the AUV in strong disturbance condition. Firstly, an observer is designed to filter out the disturbance from the output signals, and make them unable to enter the control loop. Secondly, a local approximating radial basis neural network （RBNN） is designed to compensate the nonlinear structure which is composed by uncertain hydrodynamic coefficients and coupled velocities of surge, sway, and the yaw angle. Last, the stability of the closed-loop system will be proved by Lyapunov theory. Simulation experiment results show that the proposed controller has good adaptability for viscous damping coefficients and unknown nonlinear structure and strong robustness for uncertain additional quality and changing disturbance.

关 键 词：水下机器人 观测器 径向基神经网络 自适应输出反馈 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]