Intelligent ultrasound processing applied to insulation pollution estimation

This paper presents field results for a pollution estimation system based on ultrasound noise and Statistical Auto-Associative Artificial Neural Networks (SA³N²). The system extracts spectral information from the ultrasonic noise emitted by the corona discharges that occur nearby electric insulation, then correlates this information to a previously known pollution intensity situation. The entire acquisition is performed meters away from the energized circuit. The audio is processed with the Spectral Significance Mapping (SSM) algorithm, which performs an intelligent spectral delineation and compression. The results show that the method is reliable, despite suffering the influence of moisture, since it changes the ultrasound spectrum. This effect can be minimized if the database that is used to train the SA³N² is sufficiently diversified. Due to the SA³N² generalization capacity, even new situations can be relatively well classified.