G.V. Shadsky – Dr. Sc. (Eng.), Professor, Tula State University
I.N. Isakov – Leading Process Engineer, JSC «KBP named after Academician A. Shipunov» (Tula); Post-graduate Student, Department «Technology of mechanical engineering», Tula State University
This paper considers the problem of determining the startup sequence of parts in machining on sliding head machines, CNC to minimize the time of changeover. To solve this problem was proposed a matrix method of the description of technological process (TP) treatment. While the TP pre-provided set of features-characteristics recorded in the form of matrix-rows. The work considers existing methods of cluster analysis. Determines the startup sequence of parts using different methods and metrics, build the corresponding dendrograms, counted the number of changeovers and total time. The choice of the most suitable methods and metrics were determined based on the minimum total time of changeover. Considered an option in which it is possible to choose the most appropriate method of cluster analysis, without counting the number of changeovers. To do this, the resulting dendrograms is represented as a function f(x), where x – the order details, f(x) – distance association details. Then, a dendrograms can be approximated by an integral curve and the area under the curve can be interpreted as the quantity of setups when using the given method and metric. The described approach allows to re-duce preparation-closing time processing and to reduce the minimum lot size of workpieces for a selected class of machines.
- Kim Dzh. O., M'yuller Ch.U., Klekka U.R., Oldenderfer M.S., Bleshfild R.K. Faktorny'j, diskriminantny'j i klasterny'j analiz: Per. s angl./ Pod red. I.S. Enyukova. M.: Finansy' i statistika. 1989. 215 s.
- Mandel' I.D. Klasterny'j analiz. M.: Finansy' i statistika. 1988. 176 s.