气动肌肉的轴向和径向冲击检测与区分

冲击检测是仿生驱动器研究难点。基于管腔效应，对比分析气动肌肉轴向和径向冲击特性，采用流体阻抗法建立差压信号模型；搭建冲击测试平台，通过实验研究了负载、气压、冲击强度和径向冲击位置对差压信号的影响；对比了径向和轴向冲击时差压信号相频曲线的周期性变化特性；设计自相关函数提取差压信号的周期性特征。实验结果表明，差压信号自相关函数法可有效检测与区分轴向和径向冲击，40组验证实验数据的区分准确率为97.5%. 冲击检测是仿生驱动器研究难点。基于管腔效应，对比分析气动肌肉轴向和径向冲击特性，采用流体阻抗法建立差压信号模型；搭建冲击测试平台，通过实验研究了负载、气压、冲击强度和径向冲击位置对差压信号的影响；对比了径向和轴向冲击时差压信号相频曲线的周期性变化特性；设计自相关函数提取差压信号的周期性特征。实验结果表明，差压信号自相关函数法可有效检测与区分轴向和径向冲击，40组验证实验数据的区分准确率为97.5%.

Detection and Identification of Axial and Radial Impacts on Pneumatic Muscle

Impact detection is a challenge in research of bionic driver. Axial and radial impact characteristics of pneumatic muscle are analyzed based on the tube cavity effect, and the differential pressure signal is modeled using the fluid impedance method. An impact test platform is established and used to gain the effects of load, pressure, impact intensity and radial impact position on the differential pressure signal. The periodic changes in the phase frequency curves of differential pressure signal under axial and radial impacts are compared. Autocorrelation function method is used to extract the periodic characteristics of the differential pressure signal. Forty groups of validation experiments are implemented. The experimental results show that the axial and radial impacts can be detected and identified by the autocorrelation function method of differential pressure signal. The identification accuracy is 97.5%.