基于BP神经网络的低通滤波器故障诊断.docVIP

PAGE XLIV PAGE 25 基于BP神经网络的低通滤波器故障诊断 摘 要 随着电子工业的发展，电子设备越来越复杂，传统的人工故障诊断方法已经无法满足要求，电路的故障诊断和神经网络成为当今学术界的两大热点问题。近年来，神经网络在故障诊断发面得到了越来越多的应用，神经网络故障诊断是今后故障诊断的发展趋势。 在简要的研究了低通滤波器和BP神经网络的基本原理基础上，详细说明了基于BP神经网络算法的低通滤波器故障诊断方法和设计步骤。熟悉低通滤波器的作用、类型和分类，分析哪些因素的变化可能引起低通滤波器发生故障，对低通滤波器发生的故障进行分析，确定故障类型，并对故障数据进行整理和计算。用神经网络的方法建立低通滤波器的故障诊断系统，设计神经网络故障诊断系统的结构、具体层数、输入、输出的节点数及每层的函数。利用神经网络建立过程数据与故障类型之间的对应关系，辨识出系统的正常运行状态与故障运行状态。 用MATLAB对过程进行编程仿真，并在实验室进行调试，得到的误差变化曲线基本上符合提出的期望。结果表明BP神经网络算法能够有效的应用于低通滤波器故障诊断。 关键词:BP神经网络；故障诊断；低通滤波器 Fault Diagnosis of Low-pass Filter Based on BP Neural Network Abstract With the development of the electronics industry, electronic equipment become more and more complex, the traditional manual fault diagnosis methods have been unable to meet the requirements, circuit fault diagnosis and neural networks as the two hot issues of todays academia. In recent years, the neural network in fault diagnosis baking more and more applications, neural network fault diagnosis is the development trend of the fault diagnosis in the future. After a brief introduction to the low-pass filter and the BP neural network based on the principle, detailed description of the diagnostic methods and design steps of the BP neural network algorithm-based low-pass filter failure. Familiar with the role of the low-pass filter, type and classification, and analysis of the factors which change is likely to cause the failure of low-pass filter, failures of the low-pass filter to determine the type of fault, and make fault data collation and calculation. Fault diagnosis system using neural network approach to the establishment of a low-pass filter, the structure of the designed neural network fault diagnosis system, the specific number of layers, input, output nodes and the function of each layer. Using neural network to establish the correspondence between the process data and the type of fault, identify the systems normal operation and fault operating status. Process using MATLAB programming simu