In robotics, bioinspiration brings scenarios where soft robots acquire life-like abilities and better integrate themselves into the natural world. Because of their visual and functional similarity to biological organisms, soft robots are commonly presumed to be more “natural” and therefore more appealing than traditional rigid robots. The bio-inspired soft bending actuators play a crucial role in soft robotics and have been extensively investigated at present. The existing bending actuators are mainly achieved by the bladder’s asymmetric deformation, such as pneumatic network (Pneu-Nets) bending actuators, fiber reinforcement actuators, and braid sleeve actuators. To improve the bending efficiency and strength, a bio-inspired dual-mode actuator (DMA) with a build-in angle sensor was proposed. It couples asymmetric deformation and tendon-driven through a novel fiber winding method. A soft resistive sensor was designed and prepared for easy deployment. We then fabricate three types of bending actuators and experimentally validate the bending efficiency of the DMAs. A theoretical model for the relationship between actuation pressure and bending angle is established. To evaluate its performance, the curvature, bending angle, output force, and the bending angle of the sensor feedback were measured. The results show that DMA has good bending efficiency and output force, and the tendon has an obvious reinforcing effect on the actuator. In addition, the embedded sensor has excellent embeddable deployment performance and compatibility. Finally, demonstrates the grabbing and holding of various objects using grippers composed of different lengths and numbers of DMAs. And showed the underwater flexible manipulator operation with the three-finger gripper with bending feedback as it can estimate the size of the objected like human hand.

在机器人技术中，生物灵感带来了软机器人获得栩栩如生的能力并更好地融入自然世界的场景。由于与生物有机体的视觉和功能相似，软机器人通常被认为更“自然”，因此比传统的刚性机器人更具吸引力。仿生软弯曲致动器在软机器人技术中发挥着至关重要的作用，目前已得到广泛研究。现有的弯曲致动器主要是通过气囊的非对称变形来实现的，如气动网络（Pneu-Nets）弯曲致动器、纤维增强致动器和编织套筒致动器等。为了提高弯曲效率和强度，提出了一种具有内置角度传感器的仿生双模致动器（DMA）。它通过一种新颖的纤维缠绕方法将不对称变形和肌腱驱动耦合在一起。设计和准备了一个软电阻传感器，以便于部署。然后，我们制造了三种类型的弯曲执行器，并通过实验验证了 DMA 的弯曲效率。建立了驱动压力与弯曲角度关系的理论模型。为了评估其性能，测量了传感器反馈的曲率、弯曲角度、输出力和弯曲角度。结果表明，DMA具有良好的弯曲效率和输出力，肌腱对执行器有明显的增强作用。此外，嵌入式传感器具有出色的嵌入式部署性能和兼容性。最后，演示了使用由不同长度和数量的 DMA 组成的夹具抓取和握住各种物体。并展示了具有弯曲反馈的三指夹持器的水下柔性机械手操作，因为它可以像人手一样估计物体的大小。