Welcome to the dlp wiki! This Wiki explains the the setup procedure and how to launch the included examples.

References

DLP is an implementation of my research work in the following papers:

• A Distributed Framework for Real Time Path Planning in Practical Multi-agent Systems, IFAC 2017
• Distributed Real Time Control of Multiple UAVs in Adversarial Environment: Algorithm and Flight Testing Results, accepted in ICRA 2018.

Overview

The work is about the design of a distributed algorithm for a cooperative team of agents taking local decisions for intercepting another team of intruders, in real time. For time critical applications, urgency of making real-time meaningful decisions is favoured over optimality. The setup assumes no central coordinator, and team agents take local decisions based on peer-to--peer interactions. The algorithm is based on the solution of a Linear Program in its core, in a distributed way. Hence, the name DLP. More details can be found in the above papers.

Videos

TBD

What does this package do?

This package provides two main things.

• a stand-alone C++ implementation of the proposed algorithm in dlp.cpp. A set of test files can be found in the tests folder

• The package is also organized as a ROS package. A ROS interface is provided. The main ROS nodes which interfaces the algorithm to ROS are implemented in dlp_node.cpp and dlp_node_attacking.cpp in the src folder. The ROS implementation is only done for quadrotors setup that is based on the PX4 firmware. Other ROS nodes that facilitates the interfacing are implemented in Python in the scripts folder.

Contact

• Mohamed Abdelkader: [email protected]