毕业设计（论文）-基于BP神经网络的电路故障诊断.doc

PAGE 44 PAGE 43 基于神经网络的电路故障诊断 摘 要 电路的故障诊断和神经网络是当今学术界的两大热点问题。本文主要是以模拟电路的故障诊断为例进行研究。目的在于将模拟电路故障诊断与神经网络方面的最新成果相结合，探索解决模拟电路故障诊断的一条新的途径。 在简要介绍标准BP神经网络基本原理的基础上，详细说明了基于改进BP神经网络算法的模拟电路故障诊断方法和设计步骤，根据网络总误差来自适应调节网络的学习率，加速网络的收敛过程。此算法应用于电路的故障诊断，能够对被测电路的故障进行有效并且精确的分类。以折线式有效值测量电路为例，设计了基于改进BP神经网络算法的模拟电路故障诊断系统，以实际测试数据作为训练样本进行学习训练后，对其它实际测量数据进行诊断，结果正确，验证了算法的有效性。 关键词:神经网络；故障诊断；模拟电路 Fault Diagnosis of Circuits Based on Neural Network Abstract Circuit fault diagnosis and neural network are now two hot issues in the current academic. This article mainly conducts the research take the analogous circuit fault diagnosis as the example. By introducing the basic principles of standard BP neural network, a method and design steps basis on improved BP neural network algorithm is described in this paper to solve the problem of fault diagnosis for simulation circuit. The improved algorithm could adapt the learning efficiency automatically and accelerate the network convergence process according to the network total error. The fault of the measured circuit can be classified effectively and accurately owing to the algorithm application to the circuit fault diagnose. Taking the mode of available value measurement circuit of broken line as an example, we design a simulation circuit fault diagnosis system based on the improving BP neural network algorithm. After networks study and train by actual test data as training specimen, using other actual measurement data for diagnosis, we can get correct results. The results show that this method is effective. Key word: neural network; fault diagnosis; analogous circuits 目 录 TOC \o 1-3 \u 引 言 PAGEREF _Toc232498392 \h 1 第1章 绪 论 PAGEREF _Toc232498393 \h 2 1.1 诊断工程概述 PAGEREF _Toc232498394 \h 2 1.2 故障诊断机理 PAGEREF _Toc232498395 \h 2 1.3 模拟电路故障诊断的意义 PAGEREF _Toc232498396 \h 2 1.4 模拟电路故障诊断的方法 PAGEREF _Toc232498397 \h 3 1.5 神经网络故障诊断问题的提出 PAGEREF _Toc232498398 \h 4 1.6 本设计研究的内容、目的和意义 PAGEREF _Toc232498399 \h 5 第2章 人工神经网络的简介 PAGEREF _Toc2324984