A satisficing approach to free flight

Abstract

Proposed enhancements to centralized air traffic control lead to a distributed control scheme giving pilots more autonomy but requiring more support for on-board decision-making. Satisficing decision theory offers a useful framework for intelligent decision support in distributed multiagent systems. Satisficing theory uses social utility functions that enhance the ability of the agents to perform cooperatively. Simulations of air traffic demonstrate that systems of autonomous agents employing satisficing decision-making are viable, in terms of both performance and computational overhead.