Algoritma a genetik untuk masalah program bilangan bulat multi objektif
| Main Author: | Perpustakaan UGM, i-lib |
|---|---|
| Format: | Article NonPeerReviewed |
| Terbitan: |
[Yogyakarta] : Universitas Gadjah Mada
, 2003
|
| Subjects: | |
| Online Access: |
https://repository.ugm.ac.id/27178/ http://i-lib.ugm.ac.id/jurnal/download.php?dataId=10230 |
Daftar Isi:
- . ABSTRACT This research is aimed at finding solutions in multiobjective integer programming problems. The solutions for multiobjective programming problems is compromise solutions between objective functions, i.e., finding Pareto solutions. By considering the ambiguity of DM's judgments as a human, fuzzy goals for objective functions.are incorporated. After determined membership functions ji,(x) for objective functions, assume that i=1, 2, reflecting the aspiration level of the DM for each membership functions, the problems can be formulated as augmented minimax problem. By solving the above augmented minimax problem, a Pareto solution which is nearest to the reference membership levels can be obtained regardless of its uniquness. Genetic algorithm is stochastiC (random), searching method that mimic the metaphor of natural biological evolution, like natural selection, crossover, and mutation. The rule of this game is the fittest will win. In it's application, genetic algorithm just need the evaluation function of the problem which will to optimised, not the mathematically guided algorithm is needed. In this paper, it will use genetic algorithm step by using interactive satisficing method. The result of genetic algorithm programming simulation proves that the increasing of fitness value will increase the minimized objective function value and decrease the minimized objective function value. Thus, the genetic algorithm has succeeded in finding compromise solutions for the fuzzy multiobjective integer programming. Keyword : genetic algorithm, multiobjective integer programming, fuzzy multiobjective integer programming, augmented minimax problem.