Abstract
The worldwide aircraft/airspace system (AAS) is faced with a large increase in air traffic in the coming decades, yet many flights already experience delays. The AAS is comprised of many different agents, such as aircraft, airlines, and traffic control units. Technology development will make all the agents in the AAS more intelligent; hence, there will be an increasing overlap of the declarative functions of the agents. This paper describes the basis for an Intelligent Aircraft/Airspace System (IAAS) that provides improved system performance, redundancy, and safety by utilizing the overlapping capabilities of the agents. Principled Negotiation between agents allows all the agents in the system to benefit from multiple independent declarative analyses of the same situation. Multi-attribute utility theory and decision trees are used as the basis for analyzing the behavior of different types of agents. Intelligent agents are modeled as rule-based expert systems whose side-effects are the procedural and reflexive functions of the agent. Principled negotiation also is a side-effect of the expert systems' declarative functions. A hierarchical organization of agents in the IAAS is proposed to facilitate negotiation and to maintain clear lines of authority.
Original language | English (US) |
---|---|
Pages (from-to) | 177-187 |
Number of pages | 11 |
Journal | Artificial Intelligence in Engineering |
Volume | 12 |
Issue number | 3 |
DOIs | |
State | Published - Jul 1998 |
All Science Journal Classification (ASJC) codes
- General Computer Science
- General Engineering
Keywords
- Air traffic control
- Distributed artificial intelligence
- Intelligent systems
- Multi-agent systems