毕业设计（论文）-基于BP神经网络的电路故障诊断.doc

文档介绍：I 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****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****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 II Fault Diagnosis of Circuits Based on work Abstract Circuit fault diagnosis and work 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 work, a method and design steps basis on improved BP work 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 work convergence process according to work 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 work algorithm. works 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 : work; fault diagnosis; analogous circuits III 目录引言..................................................................................................................................... 1 第 1章绪论............................................................................................................................ 2 诊断工程概述...................................................................................................................... 2 故障诊断机理.....................................................................