电子系统测试性虚拟验证与综合评估技术研究

发布时间：2017-12-28 14:23

本文关键词：电子系统测试性虚拟验证与综合评估技术研究　出处：《电子科技大学》2016年硕士论文　论文类型：学位论文

【摘要】：测试性验证是决定使用方是否接受装备测试性指标的重要环节,目前对装备进行测试性验证的主要是利用实物验证方法,实物验证具有试验成本高,故障难以注入,试验周期长等特点,因此本文提出了虚拟验证的方法。论文首先优化了虚拟验证方案,然后利用故障注入系统进行虚拟故障注入并根据试验结果评估出装备测试性指标,最后设计并开发了虚拟验证软件,论文的主要研究内容如下:1.针对目前虚拟验证方案设计不明确的问题,论文首先对此进行深入研究。虚拟验证方案主要包括试验方案的设计与故障样本量的筛选,试验方案设计是在介绍以前的利用概率论进行方案设计的基础上,提出了二项分布法的优化方法,并通过计算机编程求出一些常用方案列于表中以供查询。概率论所设计的方案往往样本量庞大,为了缩小试验样本量,本文提出了一种融合研制阶段的验证信息来减小故障样本量的方法,并对比分析该方法能够明显减小故障样本量,非常适用于昂贵设备的测试性验证方案。论文对样本量的筛选所做的研究主要包括提出用平均无故障时间进行样本量分层分配工作,以及在子模块中利用故障传播概率模型进行故障模式的优选。本文所做的对虚拟验证方案的研究充分的解决了虚拟验证方案难以确定的问题。2.针对虚拟验证中模型难以确立以及故障注入风险大的问题,论文研究了相关性虚拟建模技术并由此研发了故障自动注入系统。相关性虚拟建模按照试验样本量进行故障仿真,把被测电路构成自激振荡,根据系统注入的故障拟合出被测电路在测试性设计中测点的频率-相位差特征曲线,由特征曲线的信息得出虚拟模型。故障注入系统是利用VS与Pspice混合编程实现。3.针对虚拟验证中指标评估困难与置信度不高的问题,论文研究了利用概率统计的方法来估计测试性指标,研究了融合小样子测试性实物验证的综合评估来提高置信度。估计方法主要包括点估计与区间估计,综合评估方法中利用贝叶斯公式。并提出了利用虚拟模型进行故障诊断的方法,主要研究的是把虚拟模型转换为多信号模型,然后利用基于多信号模型的故障诊断方法对系统进行故障诊断。4.设计并研发了测试性虚拟验证分析平台软件,软件以VS作为开发环境,混合Pspice与Sql Server数据库进行编程,软件实现了虚拟验证方案设计模块、故障自动注入与虚拟建模模块、参数评估与基于虚拟模型的故障诊断五个模块。
[Abstract]:Testability verification is important to determine whether or not to accept the use of equipment test index, at present the equipment testability verification is mainly uses the physical verification method, verification test with high cost, difficult to inject fault characteristics, long test cycle, so this paper puts forward a method of virtual verification. Firstly, optimize the virtual verification scheme, then the fault injection system for virtual fault injection based on the experimental results and the evaluation of testability index, finally design and develop the virtual test software, the main contents of this dissertation are as follows: 1. in view of the current virtual verification scheme design is not clear the problem, firstly, this in-depth study. The virtual screening program mainly includes the design and verification of fault sample test plan, test scheme is introduced in the previous theory by using probability based design, put forward the optimization method of two distribution method, and through computer programming for some commonly used schemes are listed in the table for query. The design scheme of probability theory often samples in order to reduce the huge amount of test sample, this paper proposes a validation of the development stage of the information fusion method to reduce the amount of fault samples, and the comparative analysis of the method can obviously reduce the fault samples, is very suitable for the test set of expensive verification plan preparation. In this paper, the research of sample size selection mainly includes the use of MTBF for sample assignment, and the use of fault propagation probability model in the sub module to optimize the failure mode. The study of virtual verification scheme in this paper fully solves the problem that the virtual verification scheme is difficult to determine. 2., aiming at the problem that the model is difficult to establish in the virtual verification and the risk of fault injection is large, the relevance virtual modeling technology is studied, and the automatic fault injection system is developed. The correlation virtual modeling is simulated according to the sample size of the test, and the self-excited oscillation of the measured circuit is made. According to the fault injected by the system, the frequency phase difference characteristic curve of the measured point in the testability design is fitted, and the virtual model is obtained from the information of the characteristic curve. The fault injection system is realized by the hybrid programming of VS and Pspice. 3., aiming at the problem of difficulty and confidence in index evaluation in virtual verification, this paper studies the use of probability and statistics to estimate testability indicators, and studies the comprehensive evaluation of small sample testable physical verification to improve confidence. The estimation method mainly includes the point estimation and the interval estimation, and the Bayesian formula is used in the comprehensive evaluation method. A method of fault diagnosis based on virtual model is put forward. The main research is to transform the virtual model into a multi signal model, and then use the multi signal model based fault diagnosis method to diagnose the system. 4. design and development of virtual test verification and analysis software platform, the software takes VS as development environment, mixed Pspice programming and Sql Server database software to realize the virtual verification scheme design module, automatic fault injection parameter module, and virtual modeling evaluation and fault diagnosis based on virtual model five modules.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TH70

【相似文献】