【摘要】 为了提升电缆线长测量精度，为输电线缆的实际铺设奠定基础，研究基于TDR技术与深度神经网络的电缆线长时差法测量技术。结合时差法与RBF深度神经网络，实现对待测电缆线起始端点的定位，通过TDR技术向所定位的起始端点位置注入脉冲电压信号，运用模拟内插值法得出入射脉冲与反射脉冲的时间间隔，运算获得脉冲的行波波速，得出待测电缆线的长度。实验结果表明，该技术可精准定位实验电缆线的起始端点位置，位置坐标的最高误差低于0.13 m,能够测量出实验电缆线的长度，所测长度与实际电缆线长度较为接近。更多还原

【Abstract】 In order to improve the measurement accuracy of cable length and lay the foundation for the actual laying of transmission cables, the time difference measurement technology of cable length based on TDR technology and deep neural network is studied. Combining the time difference method and RBF deep neural network, the location of the starting end point of the cable to be tested is realized. The pulse voltage signal is injected into the location of the starting end point through TDR technology. Then, the time interval between the incoming and outgoing pulse and the reflected pulse is obtained by analog interpolation method. The traveling wave velocity of the pulse is calculated to obtain the length of the cable to be tested. The results show that this technology can accurately locate the starting end point of the experimental cable, The maximum error of the position coordinate is less than 0.13m, and the length of the experimental cable can be measured, which is close to the actual cable length.更多还原