陸 昱羲

<jats:title>Abstract</jats:title> <jats:p>Soft actuators have demonstrated potential in minimally invasive surgery (MIS), requiring multiple degrees of freedom and stiffness modulation facilitated by pneumatic antagonistic chambers. However, the hollow central passage in these actuators for single-port surgery complicates the internal stress distribution during the inflation of multiple chambers, making stiffness modulation mechanisms difficult to understand. In this study, a finite element analysis model is developed to explore the combined effect of the internal stress distribution and duct structure on the stiffness modulation in soft actuators equipped with multiple pneumatic antagonistic chambers. A prototype is developed to cross-validate the simulation and examine the effects of various antagonistic chambers on the stiffness modulation. The findings confirm that inflating specific chambers alters the internal stress distribution of the actuator by creating bending moments that influence deformation responses to external loads. These moments impact the stiffness and structure of the hollow central duct by modifying its geometrical moment of inertia and affecting the lengths of the tensile and compressive regions during bending, further influencing the stiffness. The effects of bending moments and the geometrical moment of inertia on stiffness modulation vary across different actuator sections owing to the varying lengths of the tensile and compressive regions during bending. These stiffness modulations affect the tip bending, force, and response time. The study findings provide insights into the mechanism of stiffness modulation for the design of soft actuators for complex MIS</jats:p>

<jats:p>The simulation-to-reality (sim2real) problem is a common issue when deploying simulation-trained models to real-world scenarios, especially given the extremely high imbalance between simulation and real-world data (scarce real-world data). Although the cycle-consistent generative adversarial network (CycleGAN) has demonstrated promise in addressing some sim2real issues, it encounters limitations in situations of data imbalance due to the lower capacity of the discriminator and the indeterminacy of learned sim2real mapping. To overcome such problems, we proposed the imbalanced Sim2Real scheme (ImbalSim2Real). Differing from CycleGAN, the ImbalSim2Real scheme segments the dataset into paired and unpaired data for two-fold training. The unpaired data incorporated discriminator-enhanced samples to further squash the solution space of the discriminator, for enhancing the discriminator’s ability. For paired data, a term targeted regression loss was integrated to ensure specific and quantitative mapping and further minimize the solution space of the generator. The ImbalSim2Real scheme was validated through numerical experiments, demonstrating its superiority over conventional sim2real methods. In addition, as an application of the proposed ImbalSim2Real scheme, we designed a finger joint stiffness self-sensing framework, where the validation loss for estimating real-world finger joint stiffness was reduced by roughly 41% compared to the supervised learning method that was trained with scarce real-world data and by 56% relative to the CycleGAN trained with the imbalanced dataset. Our proposed scheme and framework have potential applicability to bio-signal estimation when facing an imbalanced sim2real problem.</jats:p>