Soft Lightweight Small-Scale Parallel Robot With High-Precision Positioning

Small-scale (from centimeters down to micrometers) parallel robots with high precision are widely utilized in various industrial and biomedical settings, while such superiorities remain challenges for soft parallel robots (SPRs). In this work, we propose an integrated design and fabrication strategy to make up a soft lightweight (3.5 g) small-scale (
$\phi 60\times$
40 mm) parallel robot based on the dielectric elastomer actuator. Then, a hybrid model is established to describe the mapping between driving space and workspace, taking advantage of the robustness and security of the model-based method and the strong nonlinear fitting ability of the data-driven neural network method. The stiffness and workspace of the robot are analyzed. The results of trajectory tracking experiments demonstrate the accuracy and robustness of the proposed hybrid model. The average positioning error of the different trajectories is 13.4–16.6
$\,\mu\mathrm{m}$
. Finally, the zebrafish embryo puncture experiment is carried out to show the ability of micromanipulation. This research paves a new avenue for designing and controlling high-positioning SPRs, which is expected to be applied in the micromanipulation field.