假肢手仿生滑移预判方法及滑觉传感器研究

To realize safe and stable grasp of unknown objects, slippage detection and feedback is essential. However, the researches on slippage sensing to date have mainly focused on slippage detection under structured environment but not unknown objects. Inspired by the human tactile perception system, this project focuses on the investigation of the slippage detection during objects grasp by first developing an innovative slippage prediction method based on the ratio between stick and slip areas with constant judgment threshold, and then a novel slippage sensor based on core-shell structured piezoelectric nanofiber arrays. More specifically, the relationship between the ratio of stick zone area to slip zone area and the ratio of tangential force to maximum static friction force will be investigated by mechanical modeling of localized slippage aiming at determining the threshold selection criteria. Sensing mechanism of the new proposed dynamic tactile sensor array constructed by core-shell structured piezoelectric nanofibers will be then analyzed and applied to the sensor design. Additionally, force controlled coaxial electrostatic spinning technology which can directly write the nanofibers on an elastic layer will be investigated and transferred into the sensor fabrication. Finally, slippage predicting and adaptive force-tracking impedance control experiments of prosthetic hands will be performed to verify the proposed slippage prediction method and developed slippage sensor.

滑觉信息是假肢手对未知物体实现安全、稳定抓取所需反馈的重要信息。目前的假肢手滑觉感知技术研究主要集中在结构化环境下的滑移检测上，无法像人手滑觉感知系统一样实现非结构化环境下未知物体抓取时的滑移预判。为此，本项目以假肢手抓取过程的滑移预判为对象，运用感知系统仿生研究方法，突破传统方法判断阈值不恒定的限制，发展基于粘/滑区域面积比的恒阈值滑移预判新方法，研制基于芯-壳结构压电纳米纤维的新型滑觉传感器。具体包括：建立假肢手与物体间的接触力学模型，探索粘/滑区域面积比随牵引/最大静摩擦力比的变化规律，确定滑移预判的阈值选择准则；利用纵横排列的芯-壳结构压电纳米纤维束构建高密度动态力传感阵列，研究敏感单元的压敏特性，提出滑觉传感器的设计方法；研究力控同轴静电纺丝直写工艺，实现芯-壳结构纤维在弹性聚合物基底上的定位沉积；完成假肢手滑移预判和力控抓取实验，验证预判方法和滑觉传感器的有效性。

滑觉信息是假肢手对未知物体实现安全、稳定抓取所需反馈的重要信息。针对目前的滑移检测方法在非结构化环境下判断阈值不恒定无法实现准确滑移预判的难题，本项目从滑移预判新方法及其检测器件两个方面进行深入探索，项目主要研究成果包括：建立了假肢手与物体的接触力学模型，揭示了接触界面的在整体滑移发生前的局部微滑动规律，得到了正向力与最迟发生滑动点位置的关系曲线，提出了整体正向力和分布式微滑动检测相结合的滑移预判新方法，克服了非结构化环境下初始抓握力变化对滑移预判准确性的影响；提出了力控同轴静电纺丝新工艺，探明了溶液质量分数、基板速度、工作电压、喷射高度等工艺参数对成丝形貌的影响，得到了最佳工艺窗口，制备了定向芯-壳结构PVDF压电纳米纤维，以此为敏感单元构建了动态力传感阵列；制备了柔性滑觉传感器，搭建了假肢手与物体接触滑移模拟实验平台和滑移预判实验平台，完成了假肢手抓取和滑移预判实验，验证了所建接触力学模型的准确性和所提滑移预判方法的有效性。本课题的研究成果不仅为具有滑觉功能的可穿戴装备的设计提供新的思路，而且为我国具有触觉功能的高端仿生智能装备的研制提供理论和技术支撑，具有重要的研究意义和应用价值。