Target Localization and Pursuit With Networked Robotic Vehicles: Theory, Simulation, and Experiments

Abstract

This paper addresses the problem of range-based simultaneous localization and pursuit (SLAP) with networked robotic vehicles from a theoretical and practical standpoint. The work presented builds upon and extends previous theoretical research by Hung, Rego, and Pascoal on the subject of range-based SLAP using multiple trackers and a cooperative distributed estimation and control (DEC) strategy. The key novel contributions of the paper are twofold:

• Event-triggered communication (ETC) mechanisms for the DEC strategy are proposed, with formal guarantees of stability of the multiple vehicle ensemble. In this approach, each tracking vehicle only communicates with its neighbors—both for target estimation and cooperative control purposes when deemed necessary—thus reducing the cost of communications.

• The paper presents experimental results from multiple field trials conducted with three autonomous marine vehicles. These trials assess the effectiveness of the proposed DEC/ETC strategy in a real-world environment.

Simulation results are also included and analyzed. For the sake of completeness, source codes and links to auxiliary materials are provided, enabling the readers to run simulations and even implement the DEC/ETC strategy using both Matlab and ROS/Gazebo platforms.

Matlab codes: https://github.com/hungrepo/slap-etc.

ROS packages: http://github.com/dsor-isr/slap.

Aerial view of field trials: http://youtu.be/4LR4WSJHyz8.