基于随机抽样一致性算法的运动标定方法

发布时间：2018-03-29 17:26

本文选题：邦定设备　切入点：运动标定　出处：《现代制造工程》2017年10期

【摘要】：在邦定设备运动标定过程中收集到的对位平台旋转中心的样本数据包含有各种干扰,以至于利用其进行推导时过程复杂、结果精度低且鲁棒性不高。为了解决以上问题,提出一种基于随机抽样一致性算法的简单方法,该方法的构建基于样本数据分布密度模型,该模型进行一定次数的迭代筛选后,其最优解集合可进行精确推导。试验表明,在使用随机选取的10组样本数据的计算中,本方法平均计算时间为697.6s,标定精度在7.2pixel以内,并且有7组数据的结果均满足10pixel的像素精度要求。整个标定结果达到国外同类型设备的先进水平,能够完全满足设备运行的要求。
[Abstract]:In order to solve the above problems, the sample data collected during the calibration of the motion of the stationary equipment contain all kinds of interference, so that the derivation process is complicated, the result is low precision and the robustness is not high. A simple method based on random sampling consistency algorithm is proposed. The method is based on sample data distribution density model. After a certain number of iterations, the optimal solution set can be derived accurately. In the calculation of 10 groups of randomly selected sample data, the average calculation time of this method is 697.6 s, and the calibration accuracy is within 7.2pixel. The results of seven groups of data all meet the pixel accuracy requirements of 10pixel, and the calibration results reach the advanced level of the same type of equipment abroad, and can fully meet the requirements of equipment operation.
【作者单位】：上海交通大学自动化系;昆山工研院新型平板显示技术中心有限公司;昆山国显光电有限公司;
【分类号】：O212;TP391.41

【相似文献】