摘要
数控机床在机测量系统测量误差的影响因素多、变化规律复杂,利用软件建模预测补偿时间长、实时性差。XY工作台是数控机床在机测量系统的重要组成部分,为了实时补偿其在不同测量速度下的定位误差,文章设计了一种基于数字信号处理器(digital signal processor,DSP)的XY工作台定位误差实时建模预测补偿系统,介绍了其建模原理和软、硬件组成,并利用双频激光干涉仪和数控XY工作台进行实验,测量了不同运动速度下的定位误差,进行了遗传算法-反向传播(genetic algorithm and back propagation,GA-BP)神经网络与BP神经网络建模效果的验证比较。实验结果表明,应用GA-BP神经网络的定位误差建模精度比BP神经网络的建模精度高。利用DSP建立的XY工作台误差实时建模预测补偿系统的每个定位误差的预测时间为29.6μs,残差分布为-0.688 190~0.512 744μm,实时性和鲁棒性好,可以用于后续的在机测量系统各组成部分误差和综合测量误差的精确补偿以及误差补偿器的实现。
There exist many factors which influence the measuring error of CNC machine tools on-machine measuring system. So the measuring error has complex variation principle. It could be predicted and compensated by the software modeling. But it requires long prediction and compensation time and has poor real-time performance. XY worktable is an important part of CNC machine tools on-machine measuring system. In order to compensate its positioning error under different measuring speeds in real time, a real-time modeling prediction and compensation system of XY worktable positioning error based on digital signal processor(DSP) was designed. Its modeling principle, software and hardware were introduced. The two-frequency laser interferometer and the numerical control XY worktable were used to measure the positioning errors under different moving velocities, which were used to build the error models. The modeling effects by using genetic algorithm and back propagation(GA-BP) and BP neural network were validated and compared. The experimental results show that the positioning error modeling accuracy by using GA-BP neural network is higher than that of BP neural network. The prediction time of each positioning error is 29.6 μs by using the real-time modeling prediction and compensation system of XY worktable error established with DSP. The residual distribution of XY worktable positioning error is-0.688 190-0.512 744 μm. So the real-time modeling prediction and compensation system established with DSP can realize the real-time compensation and has good robustness, which can be used to compensate the error of each component and the comprehensive error of on-machine measuring system precisely and to build the error compensator.
作者
杨洪涛
顾嘉辉
江磊
庞勇军
李莉
YANG Hongtao;GU Jiahui;JIANG Lei;PANG Yongjun;LI Li(School of Mechanical Engineering,Anhui University of Science and Technology,Huainan 232001,China)
出处
《合肥工业大学学报（自然科学版）》
CAS
北大核心
2020年第2期157-161,共5页
Journal of Hefei University of Technology：Natural Science
基金
国家自然科学基金资助项目(51675004)
