Proposta de uma abordagem para modelagem e simulação de eventos discretos integrada com simulação baseada em agentes : aplicação em uma linha de montagem

Abstract

In this research we propose an approach for the construction of Discrete Event Simulation (DES) models integrated with Agent Based Simulation (ABS) to improve the representation of production systems in a simulation computational environment. This approach is composed of a UML language architecture, which has a joint to integrate macro specifications (top-down approach) from the perspective of DES, with micro specifications (bottom-up approach) from the perspective of the ABS, and a guide for DES-ABS modeling and simulation. The motivation to develop this approach is the limitations and difficulties in modeling the behavior of operators in SED models and the increasing demand for methods and techniques for the development of SBA models in the context of Operations Management. For the development of this approach a Systematic Review on ABS was conducted in the context of manufacturing, logistics and supply chain. The results of Systematic Review provided the theoretical, conceptual and practical bases for developing the approach and construction of the hybrid simulation model. The modeling and simulation process was adapted to allow the insertion of the agents. The results of the modeling and simulation process show that the practice of SBA applied in the context of Operations Management still presents some operational challenges, mainly for programming of entities belonging to the DES software, which already have pre-established rules of behavior. However, the use of DES-ABS may offer greater system modeling power due to new ways of representing the behavior of entities in a model. This potential can contribute to reduce the need to perform operational simplifications in systems with a certain level of complexity, increasing the legitimacy of the results obtained in the computational model.