Framework para escalonamento distribuído de processos utilizando sistema multiagentes em sistemas de produção
AUTOR(ES)
Arnoldo Uber Junior
DATA DE PUBLICAÇÃO
2009
RESUMO
The study of techniques for processes schedulers is linked to the creation of the first operational systems (OS), with the process schedulers algorithms with and without preemption. However, the use of processes schedulers reaches other areas besides OS. It affects all the problems where there is a set of tasks to be executed and a set of executing units. The tasks final execution time is directly affected by the execution sequence that is adopted, as in the case of production systems, which need information in real time, for the execution of tasks or for the diagnosis of problems aiming fast decision making. Thus, precision and agility in processing, in changing priorities and mainly, efficiency in managing the information are needed. This work proposes a framework for processes distributed scheduler, using the agents theory and the Genetic Algorithms (GA) heuristic search technique. In the framework modeling and application the MaSE (Multi-agent System Engineering) was used, which specifies stages for analyses and project of multi agents systems. The framework development and application were integrated to a JADE (Java Agent Development Framework), using the ontology developed in the Protégé editor. For the framework validation, the study of a case was developed using the HIPS (Hybrid Intelligent Process Scheduler) framework and the results and limitations obtained in this study were compared to another process scheduler.
ASSUNTO(S)
ciência da computação ciencia da computacao escalonamento scheduling genetic algorithms algoritmos genéticos informatica sistemas multi-agentes framework (programa de computador)
Documentos Relacionados
- Um sistema multi-agentes para processamento distribuido de linguagem natural
- Sistema de escalonamento de processos baseado em casos
- Proposta de um framework para prototipagem de sistemas heurísticos multiagentes baseados em algorítmos de colônia de formigas
- Escalonamento de tarefas imprecisas em ambiente distribuído
- UM FRAMEWORK DE SISTEMAS MULTI-AGENTES PARA APLICAÇÕES RFID