An immunity approach to strategic behavioral control

Abstract

An Artificial Immune System (AIS) paradigm, which is an engineering analog to the human immune system, is adopted to deliver the performance and robustness required by a multi-agent system. AIS offers a number of profound features and solutions, including the ability to detect changes, self-organization and decentralization, to the control of a fully distributed multi-agent system. By adopting the immunity mechanisms of AIS adapted to specify and implement the behavior of each agent, a behavioral control paradigm is developed. Effective coordination and mutual understanding between agents can be achieved by adopting such a strategic behavioral control based on their corresponding behavior. Each agent is abstracted as an independent entity that carries local information, searches for solution space and exhibits robust behavior to accomplish tasks. In this article, simulations are presented with an automated intelligent system. The significance of the behavioral control paradigm and the impact of the immunity-based behaviors on the overall performance of the transport system are examined. The simulation results illustrate the importance of behavioral control and the inter-relationship of each behavior in establishing a truly automated multi-agent system for the future. © 2006 Elsevier Ltd. All rights reserved.