欠驱动水下机器人的最优等效补偿轨迹跟踪控制

Optimal equivalent compensation trajectory tracking control forunderactuated underwater robots

针对康达效应矢量推进器输出饱和以及矢量力矩耦合等固有推进特性,为其欠驱动的水下机器人提出一种最优等效补偿控制方法.以轨迹跟踪误差为基础,设计滑模控制器,并利用径向基神经网络和扩张状态观测器对执行器饱和以及外界扰动进行实时估计与补偿.在此基础上,最优等效补偿控制借助离散动力学方程构建关于机器人递推状态偏差的代价函数,通过在推进器推力可达集内寻求等效力矩使代价函数最小,从而不仅保证控制律满足推进器矢量力矩耦合特性,还能有效抑制机器人欠驱动方向的扰动.利用Lyapunov理论最终验证该控制方法在推进器自身特性约束下渐近稳定;仿真实验数据表明,欠驱动水下机器人在该方法控制下可实现较精确的轨迹跟踪任务且鲁棒性较强.

Towards the underactuated underwater robot,an optimal equivalent compensation control method is proposed with considerations of inherent characteristics of the Coanda-effect vectored thruster,such as actuator saturation and vectored moment coupling.Specifically,a sliding mode controller is designed based on trajectory tracking errors,and then radial basis function networks as well as extended state observers are built to estimate and compensate the actuator saturation and external disturbance.Then,the proposed method formulates the cost function in terms of underwater robot’s recursive state deviation by means of the discrete dynamic equation.The equivalent moment can be obtained in the thruster reachable set such that the minimum of the cost function is achieved.The obtained equivalent moment not only satisfies vectored moment coupling of the thruster,but also can eliminate the disturbance for the underactuated robot effectively.It is proved that the proposed control method is asymptotically stable according to the Lyapunov theory.Simulation results also indicate the effectiveness and robustness of the control method when the underwater robot conducts trajectory tracking tasks.