Modeling Complex Distributed Systems with Autonomous Agents Using a Membrane Computing Paradigm
Last modified: August 20, 2007
Membrane computing is an emerging field of biologically inspired computing developed initially by G. Paun, that mimics the biochemical behavior of cells. A "membrane" is a context sensitive multi-set of rules and objects that operates simultaneously with all other membranes and computing is represented by the emergent bahvior of the system state. We have built a software model of a membrane computer that operates on atomic objects and strings and applies rules in large redundant multiplicity, thereby behaving much like biochemical processing in cellular membranes. Our model, called the agent based computing machine ("ABC machine") runs on either a single machine or a Beowulf cluster of arbitrarily large size. Because the ABC machine was inspired by cellular biochemistry, itself a complex system, we believe that its use as a complex systems modeling tool is appropriate and useful. In this presentation, several examples will be provided, including a traffic network model, a nonlinear optimization problem, and an agent based battle management application. This work was funded by Air Force Research Laboratory under an SBIR contract.