Knowledge and agent-based system for decentralised scheduling in manufacturing
The aim of the research paper is to develop algorithms for manufacturers’ agents that would allow them to sequence their own operation plans and to develop a multi-agent infrastructure to allow operation pair agents to cooperatively adjust the timing of manufacturing operations. The scheduling probl...
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | article |
| Lenguaje: | EN |
| Publicado: |
Taylor & Francis Group
2019
|
| Materias: | |
| Acceso en línea: | https://doaj.org/article/bcf8251859ce437198c7ff21b6f80253 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: | The aim of the research paper is to develop algorithms for manufacturers’ agents that would allow them to sequence their own operation plans and to develop a multi-agent infrastructure to allow operation pair agents to cooperatively adjust the timing of manufacturing operations. The scheduling problem consisted of jobs with fixed process plans and of manufacturers collectively offering the necessary operations for the jobs. Manufacturer agents sequenced and pair agents timed each operation as and when required. Timing an operation triggered a cascade of conflicts along the job process plan that other pair agents would pick up on and would take action accordingly. The conventional approach performs conflict resolution in series and manufacturer agents as well as pair agents wait until they are allowed to sequence and time the next operation. The limiting assumption behind that approach was systematically removed, and the proposed approach allowed manufacturers to perform operation scheduling in parallel, cutting down tenfold on the computation time. The multi-agent infrastructure consists of the Protégé knowledge base, the Pellet semantic reasoner and the Workflows and Agent Development Environment (WADE). The case studies used were the MT6, MT10 and LA19 job shop scheduling problems; and an industrial use case was provided to give context to the manufacturing environment investigated. Although there were benefits from the decentralised manufacturing system, we noted an optimality loss of 34% on the makespans. However, for scalability, our approach showed good promise. |
|---|