PENGEMBANGAN ALGORITMA GENETIKA UNTUK PERANCANGAN LINTAS PERAKITAN DENGAN ALTERNATIF URUTAN PERAKITAN DAN PERAKITANNYA MENGGUNAKAN KOLABORASI MANUSIA-ROBOT
The competition in the manufacturing industry is getting tougher along with the shorter product life cycles. PT Mattel Indonesia, one of the manufacturers in the toy industry has implemented cobots to increase the company's productivity. Shorter product life cycles and the use of more advanc...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/68335 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | The competition in the manufacturing industry is getting tougher along with the
shorter product life cycles. PT Mattel Indonesia, one of the manufacturers in the
toy industry has implemented cobots to increase the company's productivity.
Shorter product life cycles and the use of more advanced technologies require the
production line to fit the currently developed products. ASALBP-HRC is a system
where there are alternative ways to assemble products while being done with a
different type of operator. ASALBP-HRC cases that tend to be complicated can be
solved by a metaheuristic method that can achieve feasible solutions with faster
computational time.
The metaheuristic method proposed in this research is a genetic algorithm. The
proposed Genetic algorithm contains two main procedures, such as initial solution
construction and solution improvement process. This research proposes a
procedure for initial solution construction, tournament selection, crossover, and
mutation specifically to fit the characteristics of ASABP-HRC system. The proposed
genetic algorithm is translated into Python programming language to obtain
solutions for several ASALBP-HRC cases.
Based on the results, it is shown that the proposed genetic algorithm is capable of
matching the optimal solution in several datasets with an average gap of 8,62%
while having a faster computational time with an average time gap of 54%. For the
same computational time as the analytical method, the genetic algorithm can obtain
feasible solutions with a better objective value. Based on the experimental design,
it is known that population size (P ) and the number of generations (N ) significantly
affect the performance of genetic algorithms in finding solutions.
|
|---|