No abstract

Summary In the purpose of analyzing and checking time constraints in business processes, we propose to model and simulate them in the “discrete event system specification” (DEVS) formalism. The proposed model will contain atomic DEVS models representing the activities, supervised by, other DEVS control models, and based on one to one mapping. Hence, guaranteeing a close concordance with the temporal requirements. Thus, the simulation of the model results in outputs about the successful execution regarding to the temporal constraints soundness. Therefore, the modeler will be able to predict unsuitable temporal behavior. Thus, the model responds to the lack of modeling tools that do not offer enough formal specification of time constraints at the earlier phases of design as in, the standard “business process management notation.” The proposed model handles different type of time constraints in business process such as intra‐activities, inter‐activities, and inter‐process time constraints. The proposed model is implemented into JAVADEVS using DEVS‐Suite platform.

Complex systems are made of many elementary components in interaction. To model such systems, it is generally more convenient to decompose them into subsystems that are simpler to handle. This new division is to be made in a methodical way, by identification and complete definition of the various structures, actions and interactions of those sub-systems. In this work, the decomposition of the overall system into sub-systems is based primarily on the use of the Discrete EVent System Specification (DEVS) formalism. The obtained atomic and coupled models are formally verified and validated. Then, we use the Multi-agent Development KIT (MAD-KIT) Multi Agent Systems (MAS) operational tools to implement an industrial simulator. This simulator is used by beginner operators in the petroleum field to ameliorate the process of training and learning without stopping the real processes. The advantage of this approach is its adaptability as well as its possibilities of extension (addition of new functionalities). Moreover, the decomposition into sub-systems reduces significantly the complexity of the elements being implemented and therefore, allows a great modularity and a better legibility to the system. Our work is realised in collaboration with the production department of natural gas liquefaction (GL1/K Skikda), one of the principal hydrocarbons poles from SONATRACH complex in Algeria.

No abstract

No abstract