alvgaona/task-allocation

Multi-Robot Multi-Task Allocation Simulations

Hey there, I'm Alvaro! 👋

I'm an Electronics Engineer (EE) who graduated from the University of Buenos Aires in Argentina, and I'm currently working as a Software Developer.

Interests 🤓

• Autonomous Systems 🤖
• Computer Vision 👁️
• Computer Science 🖥️
• Aerodynamics 🌬️
• Artificial Intelligence 💡

Personal Achievements ✅

• Graduated from the Faculty of Engineering at the University of Buenos Aires as an EE. 👨🏻‍🎓
• Contributor at my favorite Machine Learning framework, @pytorch. 😁
• Ex-staff of the Laboratory of Artificial Intelligence and Robotics at the University of San Andrés.

How to reach me 📫

task-allocation's Issues

Cannot control cluster orientation

The task allocation controller controls every state variable from the cluster state-space vector but the yaw angle. After introducing the yaw angle as a variable to control, the controller cannot stabilize the cluster position and orientation.

Simulate a basic example using Cluster Space Control

As discussed on Friday, 11th with Juan, we need to start working on a very basic problem using the Multi-Robot Task Allocation algorithm presented by Y. Esam et. al and the Cluster Space Control Formulation presented by I. Mas and C. A. Kitts.

A few assumptions are imposed for this problem.

1. There are enough robots to execute the desired tasks.
2. We won't consider the case where we sacrifice robots to create a cluster of robots.
3. The mission has to be designed with tasks that can be executed by either a cluster of robots or a single robot.
4. We don't consider the heterogeneity of the fleet but we do consider it between a single robot and a cluster of robots.
5. The robots are considered holonomic and only have planar dynamics.

Control variable does not work with restrictions

Given the optimisation problem, the control variable does not behave well when a restriction on it is applied.

For instance:

s.t. u >= -0.2 and u <= 0.2

This won't even iterate once and it will return a None value.