Probabilistic Search and Track with Multiple Mobile Agents

TL;DR

This paper presents a probabilistic framework using random finite sets for autonomous agents to effectively search and track multiple moving targets with uncertain states and noisy measurements in a bounded area.

Contribution

It introduces a novel decision and control framework that dynamically switches between search and track modes for multiple agents based on probabilistic target estimates.

Findings

The proposed method successfully tracks multiple targets in simulations.

The framework adapts to targets appearing and disappearing randomly.

Simulation results show improved search and tracking performance.

Abstract

In this paper we are interested in the task of searching and tracking multiple moving targets in a bounded surveillance area with a group of autonomous mobile agents. More specifically, we assume that targets can appear and disappear at random times inside the surveillance region and their positions are random and unknown. The agents have a limited sensing range, and due to sensor imperfections they receive noisy measurements from the targets. In this work we utilize the theory of random finite sets (RFS) to capture the uncertainty in the time-varying number of targets and their states and we propose a decision and control framework, in which the mode of operation (i.e. search or track) as well as the mobility control action for each agent, at each time instance, are determined so that the collective goal of searching and tracking is achieved. Extensive simulation results demonstrate the effectiveness and performance of the proposed solution.