| Jurnal Angkasa Tahun 2010 Oleh Yuliani Indrianingsih |
ALGORITMA GENETIK UNTUK MENYELESAIKAN MASALAH OPTIMASI FUNGSI BERKENDALA DENGAN PENGKODEAN BILANGAN BULAT |
| ABSTRAK |
| Function optimization problems related to the search for a set of solutions through the analysis process with several obstacles and constraints appropriate combination of optimization purposes. In many cases not easy to solve optimization problems with nonlinear objective function exactly. In search of a solution to the problem is sometimes required complex mathematical formulations utuk provide a definite solution. Optimum solution requires a long calculation process and not practical. To solve this case requires heuristic methods.. Genetic algorithms can be used to solve the optimal solution, with the process of finding the optimal number of points based on probabilistic function. If there is one criteria considered to be single-purpose optimization problem, and if more become a multi-objective optimization problem. Genetic algorithms does not require a lot of math concepts, and can treat all forms of objective function and constraints. Because of the nature of the natural, genetic algorithms can be used to find solutions without regard to the subject matter specifically. Constraints function optimization problems with integer coding, can be solved by genetic algorithms. So as to produce the value of reliability function, cost and weight of a system. From the results of the study revealed that the application of Genetic algorithm to constraints function optimization problems with integer coding,has a weakness that is probabilistic, because it is always associated with random numbers. A large population size does not ensure better fitness value, and the greater the number of generations, would produce a high fitness value. The best solution point is reached at generation 200 and generation 500, then approached a constant. Key words : Genetic Algorithms, function optimization, integer coding. |
| download |