机器人可穿戴FBG人工皮肤感知机理及信息处理方法研究

Robot artificial skin can produce tactile sensing and object protecting in human-machine cooperation safety problem method and solving technology. In this project,sensing principle and data processing method based FBG embeded in PDMS silicon rubber is studied, accordingly, a novel and practical robot artificial skin unit based on FBG embedded PDMS silicon rubber is put forward as the last research goal. Different FBG string array, FBG measuring node and skin tactile sensor force are simulated and then designed to fabricate robot tactile sensor. Mode-coupling theory and Transfer Matrix Method are mixed with strain and temperature sensing principle of fiber embedded PDMS for the purpose of mathematical model.Ansys finite element and Optiwave software are used to simulated and analyzed to optimize artificial skin structure,dimension,parameter, static and dynamic performance.FBG node and artificial skin are fabricated via twice PDMS mould technology. Dynamic array, Interpolation approximation, neural network methods of FBG sensor array demodulation are studied and mapped. Static and dynamic artifical skin experimental system of FBG array are set up. All kinds of tactile force testing and true detecting in robot arm are to be finished,meanwhile, testing data wave and force spectrum map are calculated and presented.Optical reflective spectrum of FBG array are analyzed and extracted useful features. At last, structure desgin and fabrication technology of artificial skin are optimized over and over, the tactile sensor measurement value will be given as well. A serial artificail skin desgin and fabrication method is given.This application is meaningful and significant in view of researching and marketing area. It plays an important role in low cost product of artificial skin.

机器人人工皮肤可为当前人机协作中的安全问题提供一种全方位触觉感知解决方案。本课题拟研究FBG植入PDMS硅胶的人工皮肤力感知机理与信息处理方法，通过不同FBG串阵列、FBG测点、皮肤接触力等模拟，形成一种新的机器人力触觉传感器。研究内容：拟采用FBG传输矩阵方法、耦合模理论，结合光纤植入PDMS后的应变、温度传递机理，建立数学模型；采用ANSYS有限元、Optiwave等手段，进行局部点和整体面接触力数值模拟，优化人工皮肤的结构、尺寸、技术指标及静动态性能；研究FBG传感阵列的动态矩阵、插值逼近与神经网络理论识别的映射算法，搭建FBG解调的静、动态人工皮肤实验系统，进行力触觉各种测试和机器人臂上测试、数据波形分析与力谱云图映射，分析阵列化FBG反射谱信息、提取特征量；拟给出一套人工皮肤的结构设计与制备工艺方法，为低成本人工皮肤的制作奠定基础。

触觉是机器人实现与外部环境直接作用的必需媒介，通过人工皮肤阵列感受的触觉信息，智能机器人不仅可以可靠地抓取目标物体，还可以感知目标物体的轻重、材质、形状、一系列物理特征。研究具有柔性化、阵列化、多维力检测功能的人工皮肤触觉传感器，为提升机器人拟人化、智能化程度以及解决人机协作安全问题提供了新的方法。本课题针对具有前瞻性的机器人人工皮肤光纤光栅触觉感知方法进行研究：.首先，开展基于光纤光栅的人工皮感知机理与静动态性能仿真研究。分析硅胶外部受力、温度、变形对FBG的传递与感知效应，建立光栅温度、应变感知模型，得出硅胶内植入光纤后的感知机理与分析方法。Ansys仿真模拟3*3阵列传感节点间隔距离、栅区长度、弹性体厚度等传感器结构参数与传感阵列静动态性能关系，为后续结构优化奠定基础。. 然后，对人工皮肤触觉传感单元的结构优化方法进行研究。设计了拱形桥、中间孔等多种微结构感知单元，提高了人工皮肤感知灵敏度。研究植入光栅节点的人工皮肤阵列形浇注工艺，制备可用于大面积、阵列化的光栅人工皮肤。. 其次，研究光纤式阵列化感知数据解调与接触力信息反演方法。提出了基于三点定位和神经网络的接触力、位置反演算法与力谱云图交互现实方法，有效地提高了人工皮肤力触觉测量与感知精度。. 最后，开展了可穿戴人工皮肤的接触面压滑觉、动态触觉感知研究。设计了一种指尖滑觉感知结构，提出一种光栅中心波长二阶导数识别算法，实现初始滑动识别。设计了一种光纤触须振动感知结构，研究光栅振动信号时频分析方法，实现接触纹理定性识别。为人工皮肤外部形状接触的定性识别和定量表征提供支持。. 该课题资助发表的论文共20篇，其中SCI收录期刊12篇，EI收录20篇。与课题研究紧密相关的9篇，其中SCI/EI期刊收录7篇，EI会议论文2篇。