This paper presents an adaptive backstepping controller for a mechanism of pneumatic muscle actuators via an adaptive extended state observer. A dynamic model of the mechanism is established with two unknown parameters estimated by using adaptive laws. An adaptive extended state observer is established to estimate total disturbances and states of the mechanism. Moreover, adaptive extended state observer gains are obtained by adaptive laws and parallel D-eigenvalues, whose time-varying multiplier $n$th-order derivatives are derived by tracking differentiators. Finally, a nonlinear adaptive backstepping controller is designed and the effectiveness of the proposed method is expressed by experimental results.

中文翻译：

---

通过AESO的气动肌肉执行器机构的角度跟踪自适应反步控制

本文通过自适应扩展状态观察器提出了一种用于气动肌肉执行器机制的自适应反步控制器。利用自适应律估计两个未知参数，建立了该机制的动态模型。建立自适应扩展状态观测器来估计机构的总扰动和状态。此外，自适应扩展状态观测器增益是通过自适应律和并行 D 特征值获得的，其时变乘数 n 阶导数是通过跟踪微分器导出的。最后，设计了一种非线性自适应反推控制器，并通过实验结果证明了所提方法的有效性。