Семинар за рачунарство и примењену математику, 6. новембар 2018.

• 05. Новембар, 2018
• Коментари (0)

Наредни састанак Семинара биће одржан у уторак, 6. новембра 2018. у сали 301ф Математичког института САНУ са почетком у 14:15.

Предавач: Зоран Ракићевић, University of Belgrade, Faculty of Organizational Sciences

Наслов предавања: MULTI-CRITERIA GENETIC ALGORITHM FOR SOLVING DUAL RESOURCE CONSTRAINED FLEXIBLE JOB SHOP PROBLEM

Апстракт:

This lecture at IEEE seminar refers to the application of the multi-criteria genetic algorithm on solving a job shop scheduling problem from production planning. The primary job shop (JS) is a problem of scheduling several jobs on several machines where all the jobs have different processing routes. Dual resource constrained (DRC) problem is an extension of classical JS where the job’s operations are  constrained with the availability of two types of production resources, machines and workers. This problem is frequent in the case of highly customised, small-scale and one-of-a-kind production. The proposed multi-criteria approach is known as the fast non-dominated sorting genetic algorithm - NSGAII, (introduced by Deb et al., 2002). The NSGAII is applied to dual resource constrained flexible job shop problem (DRCFJS) with two criteria function: makespan – the total length of the schedule, and just-in-time as the absolute sum of earliness and tardiness of all scheduled jobs.

Keywords: genetic algorithm, NSGAII, dual resource constrained flexible job shop, production planning.