Abstract: Aiming at the problem of designing adaptive controller for different plants for maglev ball system, a nonlinear adaptive control approach based on combination of feedback linearization and parameter identification is presented. Firstly the maglev ball system mathematical model is formulated by using the theory of state feedback exact linearization. Secondly a nonlinear controller is designed via system state feedback, and the method of controller parameter identification online is presented. The online experiments on MATLAB platform show the system using the presented approach has more advantages than backstepping sliding model control method, which can adaptively suspend different plants in equilibrium position. Furthermore it has ideal steady-state regulation performance.

Key words: Maglev levitation ball, adaptive control, feedback linearization, parameter identification

摘要： 针对磁悬浮球系统被控对象变化时控制器自适应问题，提出了一种反馈线性化和在线参数辨识相结合的非线性自适应控制方法。基于状态反馈精确线性化方法建立磁悬浮球系统的数学模型，通过状态反馈设计了一种非线性控制器，并给出了控制器参数的在线辨识方法。MATLAB平台上在线实验结果表明，与反演滑模自适应控制方法相比，提出的方法无须在平衡位置近似线性化，可以在平衡位置实现对不同对象的自适应控制，且具有理想的稳态调节性能。

关键词: 磁悬浮球, 自适应控制, 反馈线性化, 参数辨识