The application of neural networks in group technology

摘要

This paper presents a classification method based on the Kohonen network and its modifications. Theresearch results are presented in this study. The value range was chosen to optimise the netparameters and the quality classification of machine element input patterns. Parallel, 2-layers netwhich allows joining of both geometrical technological features was used in the application. Inputpatters were written in the form a raster grid, in which every raster has a defined number code.Results shows that nets analysed could be utilised in the area of classification of machine elements.