engleski

Stochastic Algorithm Application In Project Management For Limited Series Production

This paper presents a concept for possible solutions for production scheduling in a flexible manufacturing system applying LRPT, SRPT and random rule. When applying LRPT method manufacturing system is processing the job with the longest remaining processing time. Under SRPT method at every moment of time, manufacturing system is processing that job with the shortest processing time. Random rule is shared fairly among all jobs currently in the process. The observed problems were in 50 products on 20 machines, with realistic production times. In limited series production planning there is a need for taking account of extra expenses (storage, transportation, etc.), which are necessary because cost ratio for each piece is significantly increasing in comparison to series production. This paper presents Job shop planning problem applying three different methods of stochastic algorithm. Identical initial results are used for each method and then the results are displayed. In this paper are given directives for implementation of all three given methods. As one of references for result comparison, the initial sum of shortest production times is taken into account. Conducted research shows that achieving optimal results depends on structure of each working order. The biggest flexibility has shown the solution given applying random rule. But with its appliance, large restrictions represent inability to influence on structure of processing each group of products.

production planning; project management for limited series production; Applying LRPT; SRPT and random rule

nije evidentirano