基于神经网络的划片机光栅传感器补偿方法  

Compensation Method of Dicing Saw Screw Grating Sensor Based on Neural Network

作　　者：马岩[1] 白雪峰[1] 郝鑫[1] 荣宇[1] 肇群[1] 白宇[1] 

机构地区：[1]沈阳仪表科学研究院有限公司,辽宁沈阳110043

出　　处：《仪表技术与传感器》2013年第6期121-123,127,共4页Instrument Technique and Sensor

摘　　要：提出了一种针对划片机的丝杠螺距误差补偿的新方法。采用光栅传感器的技术,克服传统补偿方法对于回程补偿不够准确的缺点;采用神经网络的算法,研制了一套适合的数学模型,进行专用分析拟合。在实际划片生产装配中,使用文中的方法,降低了螺距补偿过程中丝杠重复定位误差带来的影响,达到了良好的效果。A new kind of error compensation method of the dicing saw screw pitch is proposed in the paper.To overcome the shortcomings of traditional compensation methods for return compensates,the technology of grating sensor is adopted.Based on the algorithm of the neural network,a set of suitable mathematical model for fitting analysis is developed.In the actual scribing assembly,the influence of repeat positioning error of screw during the process of pitch compensation is reduced with the method,and proper results are achieved.

关 键 词：划片机 光栅传感器 神经网络 误差补偿 

分 类 号：TH3[机械工程—机械制造及自动化]