Akademik Veri Yönetim Sistemi
8th Symposium on Languages, Applications and Technologies, SLATE 2019, Coimbra, Portekiz, 27 - 28 Haziran 2019, cilt.74
© Baris T. Tezel and Geylani Kardas.Despite the existence of many agent programming environments and platforms, the developers may still encounter difficulties on implementing Multi-agent Systems (MASs) due to the complexity of agent features and agent interactions inside the MAS organizations. Working in a higher abstraction layer and modeling agent components within a model-driven engineering (MDE) process before going into depths of MAS implementation may facilitate MAS development. Perhaps the most popular way of applying MDE for MAS is based on creating Domain-specific Modeling Languages (DSMLs) with including appropriate integrated development environments (IDEs) in which both modeling and code generation for system-to-be-developed can be performed properly. Although IDEs of these MAS DSMLs provide some sort of checks on modeled systems according to the related DSML’s syntax and semantics descriptions, currently they do not have a built-in support for debugging these MAS models. That deficiency causes the agent developers not to be sure on the correctness of the prepared MAS model at the design phase. To help filling this gap, we introduce a conceptual generic debugging framework supporting the design of agent components inside the modeling environments of MAS DSMLs. The debugging framework is composed of 4 different metamodels and a simulator. Use of the proposed framework starts with modeling a MAS using a design language and transforming design model instances to a run-time model. According to the framework, the run-time model is simulated on a built-in simulator for debugging. The framework also provides a control mechanism for the simulation in the form of a simulation environment model.