tjards Goto Github PK

followers: 21.0 following: 18.0 repos: 27.0 gists: 15.0

Type: User

Bio: Specialized in reinforcement learning applications to robot control. Currently focused on swarming and multi-agent coordination.

Location: Canada

Welcome to my GitHub

Hi. My name is Travis. This is my GitHub. My work focuses mainly on modeling and control of aerial robotics. I prefer Python for most things. I earned my PhD in Electrical and Computer Engineering from Queen's University in 2018 and am now an adjunct professor (research) at the Royal Military College of Canada.

Publication Record (Sankey chart autogenerated using sankey-pubs):

Some links:

tjards's Projects

assembly_via_q-learning

This project implements Q-learning to automatically assemble lattice structures composed of an arbitrary number of agents.

asymmetric_lattices

Implements lattice flocking for a swarm of quadcopters with asymmetric lattice parameters and visibility constraints.

avocodo

An avocado simulator.

cracking_python

(In progress). Solutions to cracking the coding interview in python.

cracking_rust

(in development). Becoming a rustacean.

double_integrator

Implementation of classical double-integrator kinematics for a particle in 3D Cartesian space.

dynamic_encirclement

(In development). This project implements an autonomous, decentralized dynamic encirclement strategy for swarms of vehicles.

finiteactionsetlearningautomata

(Complete). A general implementation of Finite Action-set Learning Automata.

flocking_network

(Complete). An implementation of flocking for a large network of aerial vehicles

flocks_mobs_and_figure_eights

Repo for videos supporting paper submission

fourier_neural_operator

Use Fourier transform to learn operators in differential equations.

hello_rusty_world

First time playing with Rust programming language

knn_for_cybersecurity

An implementation of k-NN for classifying malicious network traffic

lisanet-deep-neural-network

A simple implementation of a deep neural network inspired by our family's cat, Lisa.

multi-agent_sim

This project implements various multi-agent coordination techniques.

pinning_swarming

Adaptive Pinning Control using Network Controlability Gramian

q_learning_particle

(in development). Implementation of Neural Dyna Q-Learning. See abstract in README file.

q_learning_template

(Ready). A template for implementing Q-Learning.

quadcopter_multi

(Complete). An open-architecture multi-agent quadcopter simulator. We implement a few modern techniques for improving the performance of aerial vehicles, including reinforcement learning and shifting planar inequalities for obstacle avoidance.