Compared with general joint-link robotic systems, bio-inspired musculoskeletal robotic systems offer the advantages of higher robustness, flexibility, and redundancy. Hence, they are a promising option for the development of next-generation robots. However, theoretical analysis regarding the superiorities of musculoskeletal systems is scarce.

This study analyzes and proves the anti-interference of a musculoskeletal system both mathematically and experimentally. a) A new method to calculate the control signals of a highly nonlinear and deeply coupled musculoskeletal system is proposed, and the redundancy range is provided. The result shows that each movement of the system can be realized by multiple control signals in a subspace. b) The anti-interference to control signals caused by the nonlinear driving mechanism of muscles and the coupled feedback effect of the skeletal system are analyzed. The simulation results prove that the bio-inspired musculoskeletal robotic arm has a smaller tracking error than the joint-link robotic arm under the same perturbations of the control signals.

The theoretical analysis is consistent with the simulation results, thereby providing a theoretical reference for the application of bio-inspired musculoskeletal robotic systems.

与普通的关节链接机器人系统相比，受生物启发的肌肉骨骼机器人系统具有更高的鲁棒性，灵活性和冗余性。因此，它们是开发下一代机器人的有希望的选择。但是，关于肌肉骨骼系统优势的理论分析很少。

这项研究在数学和实验上分析并证明了肌肉骨骼系统的抗干扰性。a）提出了一种计算高度非线性和深耦合的肌肉骨骼系统控制信号的新方法，并提供了冗余范围。结果表明，系统的每个运动都可以通过子空间中的多个控制信号来实现。b）分析了由肌肉的非线性驱动机制引起的对控制信号的抗干扰以及骨骼系统的耦合反馈效应。仿真结果表明，在相同的控制信号扰动下，生物启发型肌肉骨骼机械臂的跟踪误差小于关节连杆机械臂的跟踪误差。

理论分析与仿真结果相吻合，从而为生物启发的肌肉骨骼机器人系统的应用提供了理论参考。