基于多维特征的典型电子电路故障诊断技术研究

发布时间：2018-11-24 21:10

【摘要】：随着系统集成化程度的高速发展,国防军事等关键领域内电子系统的功能愈加完善,同时也带来了系统中电路模块可及测点少、电路信息获取困难、测试激励难以施加、故障检测率低等问题。因此必须深入研究典型电子电路的故障诊断技术,利用尽量少的测点提取尽可能多的故障特征信息,提高系统的可靠性,实现系统故障及时准确的定位。本文在分析典型电子电路结构特性的基础上,开展基于多维特征的故障诊断技术的研究,进行多维故障特征的构造与融合,优化电路可测性设计,提高系统故障诊断率。主要研究内容如下:首先,研究基于多维特征的故障诊断算法。通过单一测点提取表征电路特性的多维故障特征,解决集成电路测点少、单一电压信号信息含量少的问题;对多维故障特征进行相关性分析,提取表征信号特征的主成分,简化算法的计算复杂度;对电路进行容差分析和裕度设定,从而识别并排除器件容差、外界环境等对电路特征的影响。其次,对典型电子电路进行原理和故障特性分析,根据电路输出响应的特点,应用基于多维特征的故障诊断算法,对电路进行故障特征提取与优化,建立故障-特征依赖矩阵,缩短实际故障诊断时间。计算典型电子电路的故障诊断指标,验证算法的实用性和有效性。再次,设计典型电子电路故障诊断系统。设计系统上位机和下位机,实现用户流程可控化、故障诊断过程可视化和自动化的设计;通过划分系统级和模块级两个故障诊断模式,实现系统故障快速定位;对系统各模块进行联调,保证系统功能的完备性。最后,搭建电子电路典型应用系统,应用层次化的可测性建模方法,评估系统可测性设计,完成典型电子电路的故障诊断实物验证。实验结果表明,本文提出的基于多维特征的故障诊断技术能够及时、准确的完成电路故障诊断功能。
[Abstract]:With the rapid development of system integration, the functions of electronic systems in key fields such as national defense and military are becoming more and more perfect. At the same time, it brings less circuit modules and measuring points, difficulty in obtaining circuit information, and difficult to apply test incentives. Problems such as low fault detection rate. Therefore, it is necessary to deeply study the fault diagnosis technology of typical electronic circuits, extract as many fault feature information as possible by using as few measuring points as possible, improve the reliability of the system, and realize the timely and accurate location of the system faults. On the basis of analyzing the structural characteristics of typical electronic circuits, this paper studies the fault diagnosis technology based on multidimensional features, constructs and merges multi-dimensional fault features, optimizes the design of circuit testability, and improves the fault diagnosis rate of the system. The main research contents are as follows: firstly, the fault diagnosis algorithm based on multidimensional features is studied. In order to solve the problem of less measuring points and less information content of single voltage signal, the multi-dimension fault features characterized by the characteristics of the circuit are extracted by a single measuring point. The correlation analysis of multi-dimension fault features is carried out to extract the principal components representing the signal features and simplify the computational complexity of the algorithm. The tolerance analysis and margin setting of the circuit are carried out to identify and eliminate the effects of device tolerance and external environment on the characteristics of the circuit. Secondly, the principle and fault characteristics of typical electronic circuits are analyzed. According to the characteristics of output response, fault diagnosis algorithm based on multidimensional features is applied to extract and optimize the fault features of the circuit, and the fault feature dependence matrix is established. Shorten the time of actual fault diagnosis. The fault diagnosis indexes of typical electronic circuits are calculated to verify the practicability and validity of the algorithm. Thirdly, a typical electronic circuit fault diagnosis system is designed. The design of the upper computer and the lower computer of the system realizes the design of the user flow controllable, the visualization and automation of the fault diagnosis process, and realizes the fast location of the system fault by dividing the system level and module level into two fault diagnosis modes. All modules of the system are combined and adjusted to ensure the completeness of the system functions. Finally, the typical electronic circuit application system is built, and the hierarchical testability modeling method is applied to evaluate the testability design of the system, and the physical verification of fault diagnosis of typical electronic circuit is completed. The experimental results show that the proposed fault diagnosis technology based on multidimensional features can complete the circuit fault diagnosis function in time and accurately.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN710

【参考文献】