光纤光栅波电转换的人机仿生跟随

Bionic tracing of human and robotic arm based on FBG's wavelength-electric conversion

为拓展光纤光栅在机器人智能感知领域的应用,设计了光纤光栅感知护具并研发了光纤光栅波电转换系统及其人机仿生跟随的交互操控算法。感知护具中设置有3个光纤布拉格光栅传感单元,主要包括弹性绳的拉力环。通过拉力环对光纤光栅施加预紧力,可有效避免由于衣物或皮肤聚集引起的牵引力方向偏移。光纤光栅波电转换系统对人体手臂的肘关节弯曲角度、腕关节俯仰角度、指关节开合角度进行运动姿态特征提取,分别对应控制机械臂的方向、高度和机械爪开合角度。构建了人体手臂运动角度范围和机械臂运动角度量程的对应关系,获取了检测系数和控制系数,最终通过动态实时波长与PWM控制电信号的转换推导出系统的波电转换系数。实验结果显示,在实验量程内,机械臂的方向、高度、开合控制输出的波电转换系数最大累积误差分别是0.77111%,1.99291%,0.34117%,3个控制角度的输出偏差分别为1.13560°,1.72056°,1.82657°。光纤光栅波电转换的仿生跟随方法具有良好的控制精度,实现了机械臂运动与人体手臂运动的仿生跟随,为智能机器人仿生跟随的深入研究奠定了理论基础。

To expand the application of fiber Bragg grating sensors in the field of robotic intelligent sensing,we have designed a novel FBG forearm sensing sheath.In this workstudy,a wavelength-electric conversion control system and the an algorithm forof human-computer bionic tracing were developed.There were three FBG sensing units including tension rings in the forearm sensing sheath.The tension rings made of spandex polyurethane fibers exerted a pre-tightening force on FBGs,which could effectively avoid the traction direction deviation caused by clothing or skin gathering.The joint movements caused the spandex polyurethane fibers to drive the FBGs stretch or axial contractionng axially and the wavelengths shift of FBGs'wavelengths shift.Thereafter,the wavelength-electric conversion control system transfered the motion posture features of human elbow bending angleα,wrist pitching angleβ,and knuckle joint motion angleγto the robot arm's directionθ,heightφ,and mechanical claw opening and closing angleψ,respec-tively.By deducing the relationship between the range of the human joint movement's range and the output range of the robot arm,the detection coefficient and the control coefficient were obtained.Then,the wavelength-electric conversion coefficient was deduced by converting the dynamic wavelength to the con-trol electric signal in real-time.The experimental results demonstrate that this wavelength-electric conver-sion control system can make the robotic arm trace the human elbow bending,wrist pitching,and finger joint motion with good linear responses.Within the experimental range,the maximum cumulative error of wavelength-electric conversion coefficient of direction,height,and open/close control output are 0.77111%,1.99291%,and 0.34117%,and the deviation of three control angle outputs are 1.13560°,1.72056°,and 1.82657°,respectively.The wavelength-electric conversion control system has good con-trol accuracy and it can make the robotic arm track human forearm movements.The research results can provide a theoretical basis to further develop the FBG sensing system of bionic tracing for the robotic arm.