The Minimum Gradient Complexity Control Applied to Sensitivity Extraction of Electromagnetic Devices

This paper applies the parallel layer perceptron network trained with the minimum gradient method (PLP-MGM) to the problem of indirect sensitivity extraction of electromagnetic devices. The networks trained with the MGM are less dependent of users' defined parameters, as, for instance, the number of neurons. Some results are presented considering the indirect sensitivity extraction of a loudspeaker magnet assembly unit and an inverse scattering problem, and they show the effectiveness of the proposed approach.