基于神经网络的模拟PCB诊断方法及其智能测试系统的研究
首发时间:2010-01-13
摘要:本文提出了一种小波变换、遗传算法与神经网络相结合的模拟PCB测试新方法及其软硬件的实现。这种测试方法使用小波作为消噪工具,对信号进行消噪和小波多尺度分解,进行归一化处理后提取故障特征信息,作为神经网络的输入样本。而为了解决BP网络权值随机初始化带来的问题,用遗传算法来优化BP网络的结构和初始权值分布。在充分考虑PCB自动测试特点基础上,提出了基于DSP控制的PCB测试系统。文中研究了其故障特征提取、样本选择、诊断系统结构及其软硬件实现方法,并通过电路诊断实例,阐述了该方法的具体实现,验证了所提方法的鲁棒性。
For information in English, please click here
Study on the Neural Network Method of Analog PCB Diagnosis and Its Intelligent Testing System
Abstract:Based on wavelet decomposition, genetic algorithm (GA) and neural networks (NNs), a new testing system for Printed circuit board (PCB) is proposed. The proposed method is based on wavelet transformation that allows using the wavelet decomposition as a de-noise tool, the feature information is extracted by wavelet de-noising, multi-resolution, orthogonalization and normalization. And the input patterns are satisfied when the feature information applied to the neural networks. The proposed approach selects GA to optimize the structure and original weight distribution of BP networks. Under considering the characteristics of the printed circuit board, the controlled PCB testing system based on digital signal processor (DSP) is performed. Finally, the realization of the proposed strategy is expounded by using a practical circuit.
Keywords: BP neural network PCB testing sytem DSP wavelet transformation GA
论文图表:
引用
No.3890649235412633****
同行评议
共计0人参与
勘误表
基于神经网络的模拟PCB诊断方法及其智能测试系统的研究
评论
全部评论0/1000