Name
Title | ||
|---|---|---|
Conceptual Approach for Optimizing Air-to-Air Missile Guidance to Enable Valid Decision-making |
Abstract | ||
|---|---|---|
In this paper, we briefly introduce a concept on how the workflow of a pilot in a beyond visual range mission can be divided into different tasks in order to mimic the workflow in the behavioural control of adversary computer generated forces in training simulations. An essential part of fighter pilots' workflow is the decision making process, in which they must weigh opportunities against risks. Particularly in the weapon delivery task, valid data are a basic prerequisite for making a confident decision when weighing one's opportunities against potential risks. Concerning the applicability of artificial intelligence methods, the optimization of a missile's trajectory is used as an example to examine methods that allow an estimation of one's chances based on valid data to enable valid decision-making. For this purpose, we briefly introduce methods of optimal control and in particular deep reinforcement learning. In the future, we intend to use data generated by optimal control to validate the data provided by deep reinforcement learning methods as a basis for explainable decision-making in training simulation and threat analysis. |
Year | DOI | Venue |
|---|---|---|
2022 | 10.5220/0011302800003274 | PROCEEDINGS OF THE 12TH INTERNATIONAL CONFERENCE ON SIMULATION AND MODELING METHODOLOGIES, TECHNOLOGIES AND APPLICATIONS (SIMULTECH) |
Keywords | DocType | Citations |
Computer Generated Forces, Missile Guidance, Deep Reinforcement Learning, Optimal Control | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 3 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Philippe Ruther | 1 | 0 | 0.34 |
| Michael Strohal | 2 | 0 | 0.68 |
| Peter Stuetz | 3 | 0 | 0.68 |