基于电流的芯片级集成电路故障诊断方法研究

随着芯片级集成电路规模的逐渐增大,电路结构越来越复杂,当前故障诊断方法利用电路状态对电路故障进行检测,检测精度低。为此,提出一种新的基于电流的芯片级集成电路故障诊断方法。选择动态电流对芯片级集成电路故障进行诊断,通过Haar小波函数对芯片级集成电路进行预处理。介绍了多重分形理论基础,给出动态电流多重分形谱的计算方法。针对正常芯片级集成电路的动态电流信号求出其多重分形谱,选择一组测试向量对待测芯片级集成电路进行动态电流检测,对得到的数据进行小波变换处理,求出不同尺度下动态电流小波系数的模极大值。依据小波系数模极大值求出多重分形谱,通过其和正常电路多重分形谱之间的差异判断该电路是否存在故障。实验结果表明,所提方法诊断精度高。

Chip level integrated circuit fault diagnosis method based on the current research

With the gradual increase of the chip scale integrated circuit, the circuit structure is more and more complex. The current fault diagnosis method uses the circuit state to detect the circuit fault, and the detection precision is low. For this reason, a new fault diagnosis method based on current is proposed. The dynamic current is used to diagnose the fault of the chip integrated circuit, and the chip is integrated with the Haar wavelet function. In this paper, the basic theory of multifractal is introduced, and the calculation method of dynamic current multifractal spectrum is given. According to the dynamic current normal signal chip level integrated circuit to calculate its multifractal spectrum, select a set of test vectors to test the integrated circuit chip level dynamic current detection of the data obtained by the wavelet transform, calculated the dynamic current of the wavelet coefficients modulus maxima. According to the modulus maxima of wavelet coefficients, the multifractal spectrum is obtained, and the fault of the circuit is judged by the difference between the multifractal spectrum and the normal circuit. The experimental results show that the proposed method has high diagnostic accuracy.