Abstract
Survivability of multi-agent systems is a critical problem. Real-life systems are constantly subject to environmental stresses. These include scalability, robustness and security stresses. It is important that a multi-agent system adapts itself to varying stresses and still operates within acceptable performance regions. Such an adaptivity comprises of identifying the state of the agents, relating them to stress situations, and then invoking control roles (policies). In this paper, we study a supply chain planning implemented in COUGAAR (Cognitive Agent Architecture) developed by DARPA (Defense Advanced Research Project Agency), and develop a methodology to identify behavior parameters, and relate those parameters to stress situations. Experimentally we verify the proposed method.
Original language | English (US) |
---|---|
Pages | 555-560 |
Number of pages | 6 |
State | Published - 2002 |
Externally published | Yes |
Event | Proceedings of the Artificial Neutral Networks in Engineering Conference:Smart Engineering System Design - St. Louis, MO, United States Duration: Nov 10 2002 → Nov 13 2002 |
Conference
Conference | Proceedings of the Artificial Neutral Networks in Engineering Conference:Smart Engineering System Design |
---|---|
Country/Territory | United States |
City | St. Louis, MO |
Period | 11/10/02 → 11/13/02 |
ASJC Scopus subject areas
- Software