基于微分流形理论的虚拟仪器测量不确定度评估方法研究

There are three bottlenecks with resptect to the problems of measurement uncertaity of a virtual instrument(VI):(1)hitherto, there are no effective approaches to evaluate the measurement uncertainty of a software(algorithms), which is the kernel of a VI;(2) and so far,there are also no acceptable methods to evaluate the measurement uncertainty of a VI under static measurement occasion;and (3) up to now, there are still no acceptable methods to evaluate the measurement uncertainty of a VI under dynamic measurement occasion. Try to solve these problems, a novel mathematical model for evaluating the measurement uncertainty of the software(algorithm)adopted by a VI will be established by applying the theory of black box and that of differential manifold;then the universal distribution model will also be put forward based on exponentiale family of distribution by application of theory related to differential manifold, and this model could be applied to evaluating the measurement uncertainty of the hardware of a VI. Meanwhile, a nother mathematical model will also configurated for evaluating the measurement uncertainty of a VI under static measurement occasion by applying the theory of differential manifold and that of Maximum entropy;moreover, a nother novel mathematical model will also be presented to evaluate the measurement uncertaingy of a VI under dynamic measurement occasion by application of the theory of differential manifold and that of information fusion.Finally, an application software to evaluate the measrement uncertainty of a VI under direct measurment will be developed. All the approaches above could be applied to evaluating the measurement uncertainty of tradional instruments.

虚拟仪器测量不确定度评估这一关键基础理论问题的研究目前遇到三个瓶颈：(1)软件（算法）是虚拟仪器的核心，但目前没有公认而有效的方法评估其测量不确定度；(2)对于静态测量，在虚拟仪器各个环节测量不确定度已知的情况下尚没用有效的方法评估仪器的测量不确定度；(3)对于动态测量，由于虚拟仪器各个环节的不确定度呈现时变性，使得仪器的测量不确定度呈现时变性和高度非线性而难以确定。本课题基于黑箱理论及流形理论建立了软件（算法）不确定度评估模型；然后应用微分流形理论并基于指数族分布规律建立"统一分布规律"模型，以解决仪器单一硬件环节的测量不确定度评估问题；之后应用微分流形及最大熵原理建立虚拟仪器静态测量不确定度评估模型；最后，应用微分流形及信息融合理论建立虚拟仪器动态测量不确定度评估模型。基于上述研究，开发虚拟仪器测量不确定度评估应用程序。上述这些方法也为传统仪器的测量不确定度评估提供指导。

系统的研究了虚拟仪器中的传感器、信号调理和A/D转换三个主要的环节的测量不确定度影响因素及其分布规律，并建立了三个环节的不确定度评估模型；重点研究虚拟仪器的关键环节-软件(算法)测量不确定度的影响因素以及其分布规律，建立了软件(算法)的测量不确定度评估模型。以这些研究为基础，研究并提出了虚拟仪器静态测量不确定度评估模型。针对虚拟仪器动态测量，根据信号在测量过程中的传递流程，研究被测信号经过传感器、信号调理、A/D转换以及软件(算法)后，每一环节的测量不确定度对仪器测量结果的不确定度的影响因素以及评估模型。这些研究，为解决了软件（算法）的测量不确定评估、虚拟仪器静态测量不确定度评估以及虚拟仪器动态测量不确定度评估这三个虚拟仪器测量领域当面临的三个瓶颈问题提供了一种思路和方法。这些方法使得虚拟仪器在设计阶段而不是在仪器开发完成后就可以对仪器的测量不确定度进行评估，从而为传感器、信号调理、A/D转换以及算法的选择提供了指导。