Methods of Artificial Intelligence for Signal Parameterisation Used in the Diagnos of Technical and Biological Systems

摘要

The subject of the present paper is a proposed construction methodology for vibroacoustic signal parametrization systems based on application of artificial intelligence methods and techniques. The present considerations have been restrained to systems carrying out an automated evaluation of a selected class of vibroacoustic signals, required for diagnostics of technical and medical objects. The authors of the paper, on the basis of their multiple previous works solving problems of diagnostics of particular systems, aggregate the experience and propose methods which combine high numeric efficiency with good properties regarding the problem solving efficiency in the field of medical and technological diagnosis.