BP网络算法在平面运动测量误差补偿中的应用.pdfVIP

Vo 1. 38 No.l 第38 卷第1 期 沈阳工业大学学报 2016 年 1 月 Joumal of Shenyang University of Technology Jan. 2 0 1 6 doi: 10. 7688/j. issn. 1 仪)() - 1646. 2016. 01. 15 BP 网络算法在平面运动测量误差补偿中的应用* 哈乐，房立金 (东北大学机械工程与自动化学院，沈阳 110819 ) 摘 要:在航天对接试验平台开发项目中，针对平面运动物体非接触高精度测量的难点，提出一种 针对物体三自由度平面运动的创新性测量方法.利用激光测距传感器的并联组合，实现对物体位置 和姿态的非接触式测量.以传感器测量值为输入样本，运用 BP 神经网络算法建立误差模型，通过调 整权值因子使建立的网络模型能够拟合出测量误差，从而修正测量的数学模型.基于BP 神经网络的 误差补偿方法对平面运动物体测量误差的补偿具有明显的效果，能够满足位姿测量的精度要求. 关键词:并联测量;激光测距;非接触;传感器; BP 神经网络算法;误差模型;权值因子; 误差补偿 中圈分类号: TP212 文献标志码:A 文章编号: 1α)() -1ω6(2016)01 -∞86 -05 Application of BP network a1gorithm in error compensation of planar motion measurement HA Le , FANG Li才in ( School of Mechanical Engineering and Automation , Northeastem University , Shenyang 110819 , China) Abstract: In the development project of space docking test platform , an innovative meas町ement method for 由e planar movement of object with three degrees of f坦edom w出 proposed aiming at the difficulties in the non-contact high-precision meas町'ement of object in plan缸 movement. With the parallel combination of laser ranging sensors , the non-contact measurement for the position and posture of 0均ect could be rea1ized. Through taking the measured values of the sens町s as the input sample ，由e error model was established with BP neutral network algorithm. Through adjusting the weight factor ，往le network model could fit 由e measurement error. 白lereby ，也e mathematical model for measurement could be corrected. The results veri句， that 由e error compensation meth