Approximate Dynamic Programming Methods for Cooperative UAV Search

Authors:	Flint Matthew, BAE Systems, Advanced Information Technologies, United States Fernandez Emmanuel, University of Cincinnati, United States
Topic:	1.4 Stochastic Systems
Session:	Control, Estimation and Analysis of Stochastic Systems
Keywords:	Dynamic Programming, Stochastic Decision Processes, Modeling

Abstract

This paper considers the decentralized dynamic programming path planning decision processes of multiple cooperating autonomous aerial vehicles (UAVs) engaged in a search of an uncertain environment. However, what sets this paper apart from previous work is that a functional approximation is used for the dynamic programming (DP) cost-to-go function, resulting in improved quality of the decisions made while retaining computationally feasibility. Some of the effects of the changes are measured by simulation analysis, and the results are presented.