Loading…
Reactive Planning in Air Combat Simulation
In real-world environments such in air combat missions the agents continuously receive perceptual inputs from the environment that is highly dynamic (unpredictable) and uncertain. The aircraft often have to reorganise themselves and to make decisions under time constraints. Two modes of control are...
Saved in:
Main Author: | |
---|---|
Format: | Report |
Language: | English |
Subjects: | |
Online Access: | Request full text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | In real-world environments such in air combat missions the agents continuously receive perceptual inputs from the environment that is highly dynamic (unpredictable) and uncertain. The aircraft often have to reorganise themselves and to make decisions under time constraints. Two modes of control are thus necessary: planning and reaction. By planning we mean both building a course of action before execution, and reaction as dynamic replanning which interleaves planning and execution. Furthermore, in air combat simulation the plans adopted by the agents in response to external events are known in advance and are not generated by the agents as in other domains. To be reactive, the agents have to choose dynamically the appropriate plans and to coordinate their actions. We present how we take into account new events thanks to dynamic allocation of tasks by means of the graphs of dependencies between agents' activities. Our current work aims to extend this model of tasks and goals by integrating time notions in the selection of action plans and coordinate mechanisms. Operations on plans under time constraints are also examined in order to enable the simultaneous management of several events.
See also ADM001519. Presented at the RTO AVT Course on Intelligent Systems for Aeronautics held in Rhode-Saint-Genese, Belgium on 13-17 May 2002. Published in RTO-EN-022, p6-1-6-14, 2002. The original document contains color images. |
---|