基于在线估计的桥式吊车反演控制设计

Online-estimation-based Backstepping Control of Bridge Crane

针对桥式吊车模型不确定性问题,研究了一类基于在线估计的反演控制设计方法.相比其它常规桥式吊车控制方案,它不需要具体的系统数学模型,不需要对系统进行解耦或线性化处理,仅利用系统实时测量数据,在线估计模型信息,结合反演控制设计方法将系统分为全驱动和欠驱动两个子系统,分别进行估计和控制设计.该方法一方面能够解决吊车负载质量变化、系统摩擦力和状态耦合等模型不确定性问题;另一方面,通过李亚普诺夫函数保证了系统的控制稳定性.仿真和物理实验验证了本方法的可行性和有效性.

To deal with the uncertainty of bridge crane models,we developed an online-estimation-based backstepping control design method. Compared with conventional bridge-crane controls,the proposed method does not require a system-specific mathematical model,nor does it need to decouple or linearize the system. The proposed method uses only the real-time measurement data of the system and online estimation by the model to implement backstepping control by separating the system into fully-actuated and under-actuated subsystems.The proposed design method can handle model uncertainty due to varying crane loads,internal friction,and dynamic coupling. We confirm the stability of the closed-loop system using the Lyapunov method. The numerical simulation and experimental results demonstrate the feasibility and effectiveness of the proposed method.