中文摘要：

基于外骨骼机器人手指系统的动力学模型，提出了机器人鲁棒μ控制器的设计方法．分析了系统模型的不确定性，将负载动力学作为系统的反馈不确定，建立了合理的被控对象标称模型．基于高速DSP实现鲁棒控制器的数字化，在μ控制器的设计和数字实现过程中考虑了机器人系统的计算延时．通过实验评估采样延时对机器人控制器鲁棒性能的影响．实验结果表明，μ控制器能够抑制干扰和模型不确定性对机器人系统的影响，具有很好的鲁棒稳定性和鲁棒性能．

英文摘要：

This paper proposes a design method of the μ-synthesis robust controller for the exoskeleton robot finger based on its dynamic model. In the investigation, first, the uncertainty of the system model is analyzed, and the load dynamics is presented as a feedback uncertainty model to establish a proper nominal model of the controlled object. Then, based on the high-speed digital signal processor （DSP）, the robust digital controller is implemented, with the calculation delay of the robot system being considered in the design and implementation process. Finally, some experiments have been performed to evaluate the effect of sampling delay on the robustness of the controller. The results indicate that the proposed μ-controller effectively reduces the influences of disturbance and model uncertainty on the robot system and that it possesses good robust stability and robustness.