ВИЗНАЧЕННЯ ТЕХНОЛОГІЧНИХ НІШ КОНСТРУКЦІЙ РІЗАЛЬНОГО ІНСТРУМЕНТУ З ВИКОРИСТАННЯМ МЕРЕЖІ КОХОНЕНА

Abstract

The variety of existing tool designs makes it possible to rationally use certain types of structures depending on specific technological conditions. Relevant selection of tool designs for each specific situation determines the efficiency of the machining process. When solving the problem of selecting existing and designing new cutting tools, it is necessary to process large amounts of multidimensional information and provide decision support at each stage of design and operation of the tool. For choice or design of the tool it is necessary to investigate the influence of technological parameters on the use of the particular construction. Actual scientific researches and issues analysis is devoted to works on search of new and existing technical solutions, determination of areas for their effective use. Adaptation and use of data clustering methods in the design and selection of the tool will allow to implement them more reasonably and with less time. The purpose of the article is to automate the determination of the zones of the most effective use of a specific tool design, i.e. the allocation of technological niches for cutters using data clustering methods. A technique for automating the search for technological niches for cutting tool designs using data clustering is presented. Kohonen maps were used for this self-organizing. An example of the choice of face mill cutter designs that are most effective for specific technological conditions is given. Analysis of multidimensional data on machining use cases was performed using the analytical platform Deductor Studio. In this work, for the first time, a general approach to solving the problem of forming technological niches to determine the areas of most efficient use of appropriate types of cutting structures, substantiation of sizes of housings and plates, as well as the creation of new constructions was presented. The obtained trained Kohonen map can be used to automate the search for technological niches.