基于贝叶斯理论的ATS测量不确定度评定

Uncertainty Evaluation of Automatic Test System Based on Bayesian Information Fusion

从自动测试系统(ATS)的组成机理出发,提出一种ATS测量不确定度评定方法;其基本过程是首先基于测量信号路径,建立相应的测量链;其次,计算各个传递单元的不确定度,静态测量、动态测量分别选用贝叶斯信息融合法和贝叶斯预测法进行评定;最后,利用蒙特卡罗(MC)法计算各链的合成不确定度;通过某ATS中具体的测量链作为实例,重点分析了动态测量不确定度评定过程中遇到的不同情况及解决办法;实验表明,较其它常用评定方法,该法评定ATS静态测量得到的结果更接近理论值,不确定度变化小,评定动态测量得到结果更符合ATS动态特性且精确度高。

To assess the uncertainty of Automatic Test System （ATS）, composing mechanism of a general ATS is discussed in the first. There are three steps should be completed in the case of uncertainty evaluation of ATS.. establishing the measurement chains according to the different signal paths, evaluating the uncertainty of each transfer unit, gaining the final combined uncertainty of each chain. To complete the second step, Bayesian information fusion method and Bayesian dynamic model and forecasting method are employed in the static measurement and dynamic measurement respectively. In the last step, Monte--Carlo （MC） method is accepted to get the final results. Analyzing the solutions of the various problems in the face of dynamic measurement on the basis of a special example, and the experiment result indicates that compared with other common methods, the result got from the ATS static measurement is closer to theoretical value and its variation is less, and the result got from the ATS dynamic measurement is closer to its dynamic characteristics and it is moxe accurate.