中文摘要：

为评价大尺寸测量不确定度,特别是多站融合测量不确定度,引入蒙特卡罗仿真方法。该方法根据各传感器单元的概率特性重复采样,生成测量结果的样本,统计样本得到坐标测量不确定度,并利用计算机可视化样本的三维散点图。通过不同样本数对运算时间和准确性影响的实验,确定样本数为500。以激光跟踪仪为例进行实验,比较蒙特卡罗法、统计法和解析法三种方法得到的不确定度结果,吻合情况较好,其中与解析法比较最大偏差仅为2.7μm。实验结果表明,蒙特卡罗仿真方法可以准确评价大尺寸测量仪器及多站融合测量不确定度,融合精度优于各局部精度。

英文摘要：

To evaluate large scale measurement uncertainty, Monte Carlo simulation method was introduced especially for multi-stations fusion measurement. Repeated probability sample of each sensor unit was accomplished with this method and the measurement sample results were created and the coordinates measurement uncertainty were obtained by statistics samples so as to show 3D scatter of points through computer visualization. By affection experiment of different sample sizes on operation time and accuracy, sample size was confirmed as 500 times. Comparison results of uncertainty among Monte Carlo method, statistic and analytical method had come perfect, in which the maximum deviation was 2.7μm compared to analytical method. Experiment result showed that Monte Carlo simulation method could be used to evaluate the large-scale measurement instruments and uncertainty of multi-stations fu sion measurement and the fusion precision was superior to each local precision.