Analog Circuits Fault Detection Using Cross-Entropy Approach

作者：Li, Xifeng; Xie, Yongle^*

来源：Journal of Electronic Testing-Theory and Applications, 2013, 29(1): 115-120.

DOI：10.1007/s10836-012-5344-x

• Analog circuit
• Component tolerance
• Cross-entropy
• Fault detection
• Power spectrum

This paper presents a novel method that can detect component faults in analog circuits. Because the probability density PDF) of output voltage (current) is sensitive to the components of the circuit, the cross-entropy between the good circuit and the bad circuit is employed to detect component faults in analog circuits based on the autoregressive (AR) model. In the proposed approach, the value of each component of the circuit undertest (CUT) is varied within its tolerance limit using Monte Carlo simulation. The minimal and maximal bounds of the cross-entropy are found for fault-free circuit. While testing, the cross-entropy is obtained. If cross-entropy lies outside the tolerance limit then the CUT is declared faulty. The effectiveness of the proposed method is demonstrated via the second order Sallenkey bandpass filter circuit and continuous-time low pass state-variable filter circuit.

• 出版日期2013-2
• 单位电子科技大学