计算机工程与应用 ›› 2016, Vol. 52 ›› Issue (9): 248-252.

• 工程与应用 • 上一篇    下一篇

主成分分析和超限学习机的模拟电路故障诊断

高  坤1,2,何怡刚1,3,谭阳红1,薄祥雷1,童耀南1   

  1. 1.湖南大学 电气与信息工程学院,长沙 410082
    2.湖南科技职业学院 电子信息工程与技术系,长沙 410004
    3.合肥工业大学 电气与自动化学院,合肥 230009
  • 出版日期:2016-05-01 发布日期:2016-05-16

Fault diagnosis of analog circuits based on PCA and ELM

GAO Kun1,2, HE Yigang1,3, TAN Yanghong1, BO Xianglei 1, TONG Yaonan1   

  1. 1.School of Electrical and Information Engineering, Hunan University, Changsha 410082, China
    2.Department of Electronics and Information Engineering and Technology, Hunan Vocational College of Science and Technology, Changsha 410004, China
    3.School of Electrical Engineering and Automation, Hefei University of Technology, Hefei 230009, China
  • Online:2016-05-01 Published:2016-05-16

摘要: 为提高模拟电路故障诊断特征信息提取的完整性,实现故障模式分类的准确性,达到网络训练测试的快速性,提出了一种基于主成分分析(Principal Components Analysis,PCA)和极限学习机(ELM)相结合的模拟电路故障诊断新方法。在OrCAD16.3中通过设置仿真模拟电路元器件参数及其容差,获得电路各状态的MonteCarlo样本数据,经PCA降维提取特征信息以获得最优的特征模式,继而采用ELM对故障进行分类识别。以Sallen-Key带通滤波器电路为实例进行仿真研究,结果表明该方法具有特征提取效果好,神经网络训练学习速度快,故障诊断效率高,泛化性能好等特点。

关键词: 主成分分析, 极限学习机, 容差, 特征提取, 故障诊断

Abstract: To improve the integrality of extraction of analog circuit fault diagnosis feature information, realize the failure mode classification accurately, complete the network training and testing rapidly, this paper presents a new analog circuit fault diagnosis method based on Principal Component Analysis (PCA)and Extreme Learning Machine (ELM). The MonteCarlo sample data in each state of the circuit are got by setting parameters and tolerances components in OrCAD16.3 circuit simulation. The optimal characteristic pattern feature information is extracted by reduction dimensions using PCA. Then, the faults of the circuit are classified by ELM. As an example, the Sallen-Key band pass filter circuit is studied. The simulation results show that the proposed method has the merits of good feature extraction, fast neural network training, efficient fault diagnosis, excellent generalization performance.

Key words: Principal Component Analysis(PCA), Extreme Learning Machine(ELM), tolerance, feature extraction, fault diagnosis