Scientific journal
Scientific Review. Technical science
ISSN 2500-0799
ПИ №ФС77-57440

Personal portfolio

THE NEURAL NETWORK ANALYSIS OF THE METALS AND ALLOYS SURFACE MICROSTRUCTURE DEFECTS

Andreeva O.V. 1 Dmitriev D.V. 1
1 Nizhny Novgorod State Technical University n.a. R.E. Alekseev
The problems that hinder the widespread use of methods of the residual life prediction for the machine details and constructions, based on a quantitative assessment the degree of damage the surface microstructure of metals and alloys, were considered. The algorithm of neural method for the metals and alloys residual life prediction, using automated image processing of microstructure surface, was proposed. This algorithm is based on the singular points that were allocated from the injured surface of the microstructure. This algorithm can improve the accuracy and objectivity of the resource characteristics of metals and alloys. Also it is possible to use the resulting algorithm in conjunction with the work of experts in order to improve the quality ofa nalysis.