Reinforcement Learning Adversarial Swarm Dynamics

Abstract

Swarm robotics is a growing field that explores the implementation potential for multi-agent robotic systems completing tasks in a decentralized manor. Reinforcement Learning is a sub-field of Machine Learning that uses a feedback reward system to optimize the control laws on a given system. The Reinforcement Learning Adversarial Swarm Dynamics project will implement reinforcement learning into a simple game executed by adversarial homogeneous swarms for exploration into the feasibility and optimality of reinforcement learning in swarm robotic systems.