Multi-Agent Reinforcement Learning for Safety
Abstract
The objective of this work is to generate new fundamental science for hybrid dynamical systems that enables systematic design of algorithms using reinforcement learning (RL) techniques that are hybrid. The research proposed consists of developing novel hybrid reinforcement learning control algorithms using hybrid systems theory and validate them in experimental data-driven testbeds. The algorithms to emerge from this project will exploit real-time data exploits to robustly stabilize the system under control.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 19, 2023
- Source ID
- FA86512310004
Entities
People
- Ricardo G. Sanfelice
Organizations
- Air Force Research Laboratory
- United States Air Force
- University of California, Santa Cruz