Meu SciELO
Serviços Personalizados
Artigo
Indicadores
- Citado por SciELO
Links relacionados
- Similares em SciELO
Compartilhar
Ingeniería Industrial
versão On-line ISSN 1815-5936
Resumo
FONSECA-REYNA, Yunior César; MARTINEZ-JIMENEZ, Yailen e NOWE, Ann. Reinforcement learning applied to scheduling under real constraints. Ing. Ind. [online]. 2018, vol.39, n.1, pp. 36-45. ISSN 1815-5936.
The flow shop scheduling is a classic problem of production scheduling. The first scientific publications on production scheduling appeared more than half a century ago. However, many authors have recognized a gap between the specialized literature and the real environment problems. In this paper, the flow shop is modeled taking into account preparation time machines, setup-time between two jobs, precedence between jobs, the possibility to skip stages and as objective function minimizing the total processing time or make span. At this moment, there are no computational methods that can solve this problem taking into account all the mentioned elements. The Reinforcement Learning approach known as Q-Learning was adapted to use it in the solution of this problem. Finally, the algorithm was tested with problems of different levels of complexity in order to obtain satisfactory results in terms of solutions quality.
Palavras-chave : flow shop; optimization; q-learning; reinforcement learning.