Using artificial neural networks or Lagrange interpolation to characterize the faults in an analog circuit: an experimental study

A technique for diagnosing multiple faults in analog circuits from their impulse response function using a fault dictionary was developed. Dirac impulse input to the circuit was simulated and time domain features of the output response were used to build the dictionary. The test of a real circuit by means of a fault dictionary gives realistic results provided the simulated and measured values are similar. Consequently, the choice of the model used in the simulation is important. The precautions to realize the measurement are described