Unwrapping the influence of multiple parameters on the Magnetic Barkhausen Noise signal using self-organizing maps

The main advantage of Magnetic Barkhausen Noise as a non-destructive testing method is its sensitivity to several parameters such as microstructure, applied tension and plastic deformation. However, this noticeable property of the MBN sometimes could be a drawback. Usually, in measurements of industrial steel samples the variation of parameters occurs simultaneously. Then it is difficult to separate the influence of multiple parameters from the raw signal. This work proposes a method using trajectories traced in a type of neural network known as Self-Organizing Maps, in order to separate the influence of varying parameters on the Magnetic Barkhausen Noise row signal.