| 遗传小波神经网络的非线性模拟电路故障诊断方法 |
| |
| 引用本文： | 刘美容,何怡刚.遗传小波神经网络的非线性模拟电路故障诊断方法[J].微电子学,2009,39(6). |
| |
| 作者姓名： | 刘美容 何怡刚 |
| |
| 作者单位： | 湖南大学,电气与信息工程学院,长沙,410082 |
| |
| 基金项目： | 国家自然科学基金资助项目，教育部高等学校博士学科点专项基金资助项目，湖南省自然科学基金资助项目，国家高技术研究发展计划基金资助项目 |
| |
| 摘 要： | 提出了一种将遗传算法(GA)、神经网络与小波变换相结合对非线性模拟电路进行故障诊断的方法;分析了传统BP型神经网络在非线性模拟电路故障诊断中存在的缺陷;提出了一种新的解决方法--利用小波变换对非线性电路故障信号进行预处理,对故障信号中的冗余信息进行剔除,然后利用遗传算法优化BP网络参数,如网络权值、阈值等.利用该方法对非线性电路进行故障诊断,有利于提高神经网络对电路故障诊断的智能性及识别故障类别的能力,提高故障诊断的精度与速度.实验结果表明,该方法是可行的.
|
| 关 键 词： | 遗传算法 神经网络 小波变换 故障诊断 非线性电路 |
Research on Fault Diagnosis Method for Nonlinear Analog Circuits Based on GA, BPNN and Wavelet |
| |
| LIU Meirong,HE Yigang.Research on Fault Diagnosis Method for Nonlinear Analog Circuits Based on GA, BPNN and Wavelet[J].Microelectronics,2009,39(6). |
| |
| Authors: | LIU Meirong HE Yigang |
| |
| Abstract: | A failure diagnosis method for nonlinear analog circuits based on wavelet transform, neural network and genetic algorithm (GA) was presented.In order to improve restrained Back-Propagation Neural Network (BPNN) in application of failure diagnosis to nonlinear analog circuit, wavelet transformation was used to preprocess fault signals and distill redundant information in failure signals.Based on genetic algorithm, BP-type neural network was optimized to improve threshold and weight parameters.The new method can enhance the failure-recognition capability and intelligence for failure diagnosis, and improve its accuracy and speed.Tests results validated the proposed method. |
| |
| Keywords: | Genetic algorithm Neural network Wavelet transform Fault diagnosis Nonlinear circuit |
| 本文献已被 万方数据 等数据库收录！ |
|
