基于神经网络的模拟电路故障诊断专家系统研究基于神经网络的模拟电路故障诊断专家系统研究基于神经网络的模拟电路故障诊断专家系统研究摘 要随着电子工业的迅猛发展，模拟电路故障诊断问题已经引起广泛的关注，而且是国内外专家在设计和使用电子系统的一大难题。一些已有模拟电路的诊断方法只适用于特定条件下（如开路、短路等）的电路诊断，却很难发现由电路中的电子器件的容差变化引起的软故障。迄今为止，文献中很少对软故障即容差电路的故障诊断给出系统而有效的方法，本文将这一问题进行了研究探讨。针对传统诊断技术的局限性，讨论了利用神经网络方法诊断模拟电路软故障的方案，通过小波变换提取故障特征，并利用神经网络的非线性映射特性逼近故障诊断模型。针对传统方法的局限性，本文提出了具体的故障诊断方法，研究了基于Levenberg-Marquardt算法和动量法相结合的神经网络诊断方法，用来自模拟实验的实例对神经网络进行了训练仿真。诊断结果表明本文提出的方法是快速而有效的。此研究将为复杂模拟电路故障诊断甚至集成电路提供新的理论依据和诊断方法。此外，传统故障诊断专家系统存在不能进行自学习、自适应，知识获取困难，推理匹配冲突等不足，本文将小波分析、神经网络技术融入到专家系统的构建当中，利用小波分析提取模拟电路故障特征，用神经网络的训练代替诊断专家系统知识获取部分，用训练好的连接权和阈值代替专家系统知识部分，专家系统推理部分则通过权值数据与输入数据的运算来完成。针对一低通滤波器电路，研究开发了基于神经网络的模拟电路故障诊断专家系统。系统采用MATALB的GUI编程实现了以下功能：一是特征提取算法的实现，用户根据需要可以选择三种特征提取算法提取故障信号特征。二是神经网络参数设置，如设置隐层神经元的个数、学习率的大小等。三是神经网络的训练和故障诊断。该系统在一定程度上改善了传统专家系统不能进行自学习、自适应，知识获取困难，推理匹配冲突等不足，提高了诊断的自动化和智能化水平。关键词 模拟电路；故障诊断；神经网络；小波变换；专家系统The Study of Expert System Based on Neural Network of Analog Circuit Fault DiagnosisAbstractAlong with the rapid development of electronics industry, the analog circuit fault diagnosis question already aroused the widespread interest, moreover, it is a big difficult problem to the domestic and foreign experts when they design and use the electron system. Some analog circuit diagnosis methods which already existed are only suitable under the special condition, for example, opening, short circuit and so on. it is very difficult to discover the soft fault which causes by electronic devices tolerance change in electric circuit. Until now, very few literature gives the systematic and effective method to the soft fault, this paper research this question.In view of the traditional limitation of diagnosis technology, we discussed the plan which use neural network to diagnose soft fault of analog circuit, use the ability of wavelet transformation to extract the circuit fault character and neural network which has misalignment mapping characteristic to approaches the failure diagnosis model. In view of the limitation of traditional fault diagnosis method, this paper proposed the concrete fault diagnosis method, studied the neural network diagnosis method which the unit Levenberg-Marquardt algorithm and the momentum law, used the concrete example to neural network training simulation. The diagnosis result indicated that the method which the paper proposed is fast and effective and this research will provide the new theory basis and the diagnosis method for the complex analogous circuit failure diagnosis even integrated circuit. In addition, the tradition failure diagnosis expert system exists insufficient which cannot carry on self-study, auto-adapted, difficult to gain the knowledge, and match conflict when it inference, and so on. This article uses the wavelet analysis and neural network technology as part of the construction of expert system. Use wavelet analysis extract analog circuit fault characteristic, the expert system knowledge gaining part is replaced by neural networks training, and use the connection power and the threshold value of Back Propagation (BP) neural network which had been well trained replaces the expert system knowledge library. The expert system inference part completes through the operation of the weight data and the input data. In view of a low pass filter electric circuit, we researched and developed an expert system based on BP neural network for analog circuit. The system uses MATALB GUI programming to realize the following function, firstly, the algorithm realization of feature extraction and algorithm selection, user can choose the different diagnosis algorithm according to the users needs to realize to the fault feature extraction; secondly, set parameter of BP neural network, user can set hid level integer, study rate size and so on; thirdly, neural network training and fault diagnosis.This system makes up the flaw which traditional expert system could not self-study, the auto-adapted, overcomes the insufficiency of the traditional electric circuit fault diagnosis method, and enhances the diagnosis automation and the intellectualized level.Keywordsanalog circuit, fault diagnosis, neural network, wavelet transform, expert system不要删除行尾的分节符，此行不会被打印目录摘 要IAbstractII第1章 绪论11.1 课题研究背景、目的及意义11.2 模拟电路故障诊断的发展及研究现状21.2.1 模拟电路软故障诊断发展现状31.2.2 人工神经网络发展现状41.2.3 小波变换理论发展现状51.2.4 专家系统发展现状61.3 神经网络专家系统81.3.1 神经网络与专家系统结合的可行性81.3.2 神经网络专家系统出现的必要性91.4 模拟电路故障诊断系统的结构和功能91.5 本文的工作10第2章 人工神经网络和小波分析基本理论122.1 神经网络的基本原理122.2 BP网络的结构132.2.1 模型结构132.2.2 BP算法142.3 小波变换基本理论182.3.1 多分辨分析222.3.2 小波包分析232.4 本章小结28第3章 基于小波变换的神经网络诊断方法的研究293.1 诊断系统概述293.2 诊断系统各单元具体实现303.2.1 小波变换303.2.2 特征提取323.2.3 神经网络的设计333.3 辅助工具PSpice363.4 本章小结37第4章 故障诊断实例384.1 实验仿真与分析384.1.1 低通滤波器电路及基本诊断思路384.1.2 多分辨分析BP网络的诊断方法实现394.1.3 小波包分析BP网络的诊断方法实现454.1.4 诊断结果分析484.2 本章小结48第5章