Using nested Petri nets for modeling multiagent systems

2003

An important issue regarding the deeper understanding of social and organisational phenomena is the interaction and the linkage between micro and macro layers. The micro-macro-link deals with the emergent aggregation of processes and actions on the micro-level and with structures, rules and roles on the macro-level. This paper presents recent results of the “socionics” project ASKO, which is a fusion of sociology and computer science. Due to the interdisciplinary character of our project we use a graphical representation of our models: a special dialect of high-level Petri nets, namely reference nets, has been chosen, since it offers a simple and executable representation with a clear mathematical foundation.

Journal of Loss Prevention in the Process Industries, 2016

In the field of hazardous chemicals, there are many flammable materials located in different places within an industrial area. It is possible that major fires occur at the same time due to terrorism or vandalism (i.e. a security point of view). These fires might lead to domino effects at different locations around. Whether the fire emergency response preparedness, which is organized to fight against one of the fires, can handle the multiple fires, is studied in this article and a Petri-net based simulation approach is proposed. Through Petri-net based simulation, the process of fire-fighting can be revealed. A model of fighting against three fires is established, and the strategies of fire fighters staffing arrangements are analyzed. The results show that in most cases the distribution according to fire severity is better than the average distribution, but in some conditions the average distribution strategy prevails. Different backup staffing strategies are also compared, and the results indicate that the improvement by increasing staffing backups also depends on the fire conditions.

Procedia Technology, 2013

Simulation & Gaming, 1996

The modeling and simulation of the dynamic behavior of organizational systems is essential for understanding their current functionality and for supporting business process analysis and reengineering. The behavior of organizational systems is viewed here in terms of the behavior of their business processes. Multi-level Modified Petri nets are used as the underlying formalism for modeling and simulating organizational systems behavior at various abstraction levels. Each level of abstraction corresponds to one or more levels of the involved part of the organizational hierarchy. The constructed models represent existing or redesigned business processes. Simulation and graphical animation are used for discovering problems in existing business processes as well as in the proposed solutions. The whole approach is supported by a development environment called ERMIS.

2001

This paper presents an experience of a multiagent model formalization using coloured Petri nets, applied to the study of an hunting management system. The multiagent model of the hunting activity is presented, with simulation results. Then we describe Petri nets model assumptions and we give details about system modules. We analyze the model and we compare formal properties to multiagent simulation results. Pertinence of the approach is briefly discussed

International Journal of Advanced Computer Science and Applications, 2013

this paper deals with the modeling of interaction between agents in Multi Agents System (MAS) based on Agent Petri Nets (APN). Our models are created based on communicating agents. Indeed, an agent initiating a conversation with other can specify the interaction protocol wishes to follow. The combination of APN and FIPA Protocols schemes leads to a set of deployment formal rules for points where model interaction can be successfully implemented. We introduce some models FIPA standard protocols.

Self-organizing systems have the general property that the primitive elements that comprise the system interact in order to achieve emergent global properties. We describe an extended Petri net simulation architecture for such systems that is based on interacting agents that carry out autonomous actions, yet collaborate with each other in structured ways. There is no central controlling entity. The approach was motivated by an application to mission planning for unmanned air vehicles (UAVs), but is applicable to more general agent-oriented systems.

International Journal of Computer Applications, 2012

In this work, we propose a new formal tool called CGPN (Conceptual Graph Petri Nets) which is a combination of CG (Conceptual Graph) and CPN (Color Petri Net) to model collaborative behavior of agents in a MAS(Multi Agent System) to achieve some goals. The CG is used to represent knowledge and on the other side CPN is used to model the concurrent and dynamic aspects of a system. It is difficult to extract precise information from MAS which is dynamic in nature. Modeling MAS with CGPN will help in representing the knowledge and dynamic behavior together. Finally, the CGPN model for MAS is tested for deadlock freedom and reachability analysis to verify its correctness.

1996

This paper explores the possibility of applying distributed arti cial intelligent systems and multi-agent systems to help in the design of engineering software systems. We de ne an environment consisted basically by arti cial designers that based on their knowledge, help a human designer to obtain the design of the software systems. The software design engineering problem we take as example is the design of the interlocking software for distributed track-vehicle controllers. The design of these modules are based on Petri nets, from which we can derive invariants that are used as the basis to characterize agent's knowledge in our design aid environment.

Abstract:-This paper presents a Petri net approach to modeling communications of mobile software agents in multiagent systems. Mobile agents are conceptualized in Petri net semantics by formalizing in terms of basic agent template. The proposed software agent components constitute the elements and building blocks of the distributed mobile agent systems. A basic agent template is formalized as an entity consisting of a set of actions, a set of rules, which govern the agent communications.