ADC的测量不确定度评估方法研究

发布时间：2018-01-09 01:13

本文关键词：ADC的测量不确定度评估方法研究　出处：《陕西科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：ADC在数字化在计量测试仪器中起关键作用,它的准确性能直接影响仪器的性能。ADC性能检验的重要方法之一就是对其响应特性进行误差分析,而分析误差的现代手段就是进行不确定度评估,特别是对其动态性能进行不确定度评估。ADC的动态性能评估是目前测试和计量领域面临的难题之一,为尝试解决该问题,本文主要完成如下工作:(1)研究了测量不确定度的基本原理,分析了不同的测量不确定评估方法特点及其应用范围,提出可以利用神经网络算法进行ADC的动态性能的测量不确定度评估;(2)构建了高速ADC的动态测试平台,并对ADC的性能测试方法进行了比较和分析,包括基于直方图的静态测试方法和基于FFT的动态测试方法,最后选择FFT测试方法对高速ADC进行动态测试;基于FFT研究了ADC的五个动态性能参数:噪声系数、信纳比、有效位、总谐波失真和无杂散动态范围,由此为ADC的动态性能的测量不确定度评估奠定基础;(3)在对高速ADC的动态测试研究过程中发现可以利用噪声信号有效地提高ADC的转换性能,从而提高ADC的转换精度和抗干扰性;在测试中也获得了ADC的最佳动态性能参数,这些参数可以作为ADC动态性能的测量不确定度评估的来源;(4)针对ADC的动态测量不确定度评定这一难点问题,提出利用人工神经网络算法对ADC进行测量不确定度评估;为此建立了基于神经网络ADC测量不确定评估的数学模型,并基于该模型应用MATLAB开发了ADC动态性能评估应用程序;(5)为验证上述模型的有效性,以ADI的AD6645-105为实例进行测量不确定度评定;为此,先应用GUM中的A类和B类对其性能参数进行测量不确定度评估,再应用基于神经网络的算法对其动态性能进行不确定度评估;之后对这两种评估方法的结果进行对比分析,得到的结论是:应用神经网络算法可以更快速而准确的对ADC的动态性能参数进行评定。
[Abstract]:In the ADC in the digital measurement instruments play a key role, one of the important methods of accurate performance directly affects the performance of.ADC performance test instrument is its error analysis on the response characteristics of modern means of error analysis is the evaluation of uncertainty, especially the dynamic performance is the uncertainty evaluation of dynamic performance evaluation.ADC is one of the problems currently facing the test and measurement field, in order to try to solve this problem, this paper mainly completed the following work: (1) research on the measurement uncertainty of basic principle, analyzes the different measurement uncertainty evaluation method of characteristics and application range of measurement, put forward dynamic performance using ADC neural network algorithm the evaluation of uncertainty; (2) establishes a dynamic test platform of high speed ADC, and the ADC performance testing methods are compared and analyzed, including based on histogram The method of static testing and dynamic testing method based on FFT, the final choice of FFT test method of dynamic testing of high speed ADC; five dynamic performance parameters of FFT ADC were studied based on noise coefficient, SINAD, effective, total harmonic distortion and spurious free dynamic range, the dynamic performance of the ADC measurement uncertainty to lay the foundation of evaluation; (3) in the course of study on dynamic test of high speed ADC that can effectively improve the ADC conversion performance using the noise signal, so as to improve the anti-interference and the conversion accuracy of ADC; in the test also won the best dynamic performance parameters of the ADC, these parameters can be used to measure the dynamic performance of ADC no source of evaluation; (4) according to the uncertainty of the difficult problems in evaluation of dynamic measurement of ADC, ADC of the measurement uncertainty evaluation using artificial neural network algorithm is established based on this; Neural network ADC measurement uncertainty evaluation mathematical model, based on the model developed by MATLAB ADC dynamic performance evaluation of the application; (5) to verify the validity of the model, using ADI AD6645-105 for evaluation of measurement uncertainty for example; therefore, the first application of GUM A and B on its performance the parameters of the measurement uncertainty evaluation, and the application of neural network algorithm for uncertainty evaluation of the dynamic performance based on the two; after the assessment results were analyzed, the conclusion is: the application of neural network algorithm for the dynamic performance of the ADC parameter is more rapid and accurate assessment.

【学位授予单位】：陕西科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN792

【参考文献】