Analog Circuit Fault Diagnosis Based on Levenberg-Marquardt Learning Algorithm

Abstract:

This paper presents a fault diagnosis method of BP neural network based on Levenberg-Marquardt learning algorithm. First, the use of principal component analysis to reduce the dimension of the fault sample reduced BP neural network input variables. Then use the Levenberg-Marquardt learning algorithm to adjust the network weights. Levenberg-Marquardt learning algorithm is combination of the Gauss - Newton algorithm and steepest descent algorithm. It has Gauss - Newton algorithm of local convergence and gradient descent algorithm of the global characteristic. So it has higher convergence speed, reduces the training time, to a certain extent, overcomes the problem of traditional BP network convergence speed slow and easy to fall into local minimum point. Simulation results demonstrate the correctness and accuracy of this fault diagnosis method.

• Periodical: Applied Mechanics and Materials (Volumes 380-384)
• Pages: 979-982
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.380-384.979
• Online since: August 2013
• Authors: Huang Guo, Bao Ru Han, Guo Fang Zhang
• Keywords: Fault Diagnosis, Gradient Descent Algorithm, Levenberg-Marquardt Learning Algorithm

References

[1] Zhao Guonan, Guo Yushun . Diagnosis of Analogue Circuit Faults [M]. Ha Erbing: Ha Erbing Institute of Technology Press, (1991).

[2] Yang Shiyuan. Simulation system of fault diagnosis and reliable design [M]. Bei Jing: TsingHua University Press, (1993).

[3] Zhou Yufen, Gao Xijun. Diagnosis of Analogue Circuit Faults [M]. Bei Jing: National Defence Industry Press, (1989).

[4] Gao Hongshen, Tao Youde. Improve of BP Neural Networks Model, Systems Engineering-theory & Practice, 1996(1): 37-40.

[5] F Aminian，M Aminian, H．W．Collins. Analog Fault Diagnosis of Actual Circuits Using Neural Networks. IEEE Trans. Instrum. Meas, 2002, 51 (3) : 44-550.

DOI: 10.1109/tim.2002.1017726

[6] He Yigang, Liang Gechao, A Method for Fault Diagnosis of Analog Circuit by BP Neural Network, Journal of Hunan University(Natural Sciences. 2003, 30(5): 35-39.

[7] Hu Hui，Zhang Biao, Fault diagnosis in analog circuits based on PCA rough set and neural networks. Automation & Instrumentation, 2009(3): 35-37.

[8] Baoru Han. Based on Compact Type of Wavelet Neural Network Tolerance Analog Circuit Fault Diagnosis. ICIECS2009, 2827-2830.

[9] Wang Cheng, Chen Guangju, Xie Yongle. Fault Diagnosis in Analog Circuits Based on Principal Component Analysis and Neural Network. Journal of Electronic Measurement and Instrument. 2005, 19(5): 14~17.

DOI: 10.1109/icccas.2004.1346386