Self-Sensing Pneumatic Compressing Actuator

2020 
Using soft pneumatic actuator is a feasible solution in the complex unstructed environment, owing to their inherent compliance, light weight and safety. However, due to the limitations of soft actuators' materials and structures, they fall short of motion accuracy and load capacity, or need large-size, bulky compressors. Meanwhile, in order to gain better control, it is essential for them to sense the environments as well. This leads to high-price sensors or complicated manufacture technique. Here, a self-sensing vacuum soft actuation structure is proposed, aiming at acquiring good balance among precision, output force and actuation pressure. The actuator mainly comprises of a flexible membrane and a compression spring. When actuated, the flexible membrane outside the actuator compresses the internal spring skeleton, realizing large contractile motion in axial direction. Its built-in force sensor can indirectly measure the absolute displacement of the actuator with certain accuracy (about 5 F.S). Besides, it does not require high actuation pressure to generate enough output force. The actuator is quite easy to manufacture with low cost, and there are a variety of materials to choose from. We established quas-static models for actuators built of two different kinds of membrane materials, and tested their accuracy and output force. In addition, to break through the limits of vacuum actuation, a method of positive-negative pressure combined actuation has been proposed, which lowers the requirements for air source equipments, increases actuation pressure and reduces potential safety threats at the same time. This kind of soft actuators can also effectively resist and detect impacts. The design of a two-finger dexterous robot hand and robot joint based on this soft actuator illustrates its broad application prospects in the fields of mobile robots, wearable devices and human-robot interaction.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    78
    References
    0
    Citations
    NaN
    KQI
    []