【摘要】 拉线位移传感器将拉绳的位移转变为滑轮的转动，通过角度测量得到拉线自由端的位移，安装方式和使用环境都会对测量结果产生影响。本文基于拉线位移传感器的工作原理建立了拉线的力学模型，分析了误差影响因素，同时分析了其角度测量部件-增量式旋转编码器引入的误差。由于目前比较普遍使用的误差补偿模型优化算法难以获得较好的补偿效果，本文提出了一种基于粒子群算法(PSO)优化的BP神经网络模型的误差补偿方法。以KS120系列拉线位移传感器为研究对象，进行了实验研究，结果显示：在全量程范围内，使用该方法进行误差补偿后的拉线位移传感器精度由0.136%FS提高到0.007%FS,提高了95%。最后将本模型与基于多项式拟合算法和传统BP神经网络算法的补偿系统进行实验对比，结果显示补偿效果亦优于这两种方法。更多还原

【Abstract】 The pull wire displacement sensor converts the displacement of the rope into the rotation of the pulley, and obtains the displacement of the free end of the rope by measuring the angle. The installation method and the use environment will affect the measurement results. Based on the working principle of the displacement sensor, the mechanical model of the pull wire is established. The influence factors of the error are analyzed. At the same time, the error introduced by the incremental rotary encoder, the angle measuring part of the displacement sensor, is analyzed. Since the current common optimization algorithm of error compensation model is difficult to obtain a good compensation effect, an error compensation method is proposed based on particle swarm optimization(PSO)optimized BP neural network model. KS120 series pull wire displacement sensor is taken as the research object, and the experimental study is carried out. The results show that the accuracy of the pull wire displacement sensor is improved from 0.136% FS to 0.007% FS by using the method in the full range, which is 95% higher. Finally, the model is compared with the compensation systems based on polynomial fitting algorithm and traditional BP neural network algorithm respectively, and the results show that the compensation effect of the proposed method is better than that of these two methods.更多还原