Fair Observation of Multiple Moving Targets in Cooperative Multi-UAV Systems

Abstract

Cooperative Multi-robot observation of moving the targets (CMOMMT) represents the problem in which a group of autonomous mobile robots equipped with a limited range of sensors that can be used to keep moving targets in observation. The robots plan the movement of the target in coordination to maximize the observation time during which each target appears within the sensing range of at least one robot. We present a novel multi-objective optimization model based on UAVs for CMOMMT schemes which features fairness of observation among different targets as an additional objective. The proposed model uses a quad-tree data structure to model the movement decisions of UAVs in order to maximize duration and the variable qualities (resolutions) of observations. For any given future time, a probabilistic occupancy map for each target is estimated in a Bayesian framework. The experiment results display the effectiveness of the proposed method.