Lyapunov Design for Robust and Efficient Robotic Reinforcement Learning

TL;DR

This paper presents a Lyapunov-based cost-shaping method for reinforcement learning that significantly reduces sample complexity and enhances stability in robotic control tasks, demonstrated on hardware and simulation benchmarks.

Contribution

It introduces a novel cost-shaping technique using Control Lyapunov Functions to improve stability and reduce data requirements in robotic reinforcement learning.

Findings

Achieves stabilizing controllers with seconds to minutes of data on hardware.

Requires orders of magnitude less data than standard methods in simulations.

Provides theoretical guarantees for stability with smaller discount factors.

Abstract

Recent advances in the reinforcement learning (RL) literature have enabled roboticists to automatically train complex policies in simulated environments. However, due to the poor sample complexity of these methods, solving RL problems using real-world data remains a challenging problem. This paper introduces a novel cost-shaping method which aims to reduce the number of samples needed to learn a stabilizing controller. The method adds a term involving a Control Lyapunov Function (CLF) -- an `energy-like' function from the model-based control literature -- to typical cost formulations. Theoretical results demonstrate the new costs lead to stabilizing controllers when smaller discount factors are used, which is well-known to reduce sample complexity. Moreover, the addition of the CLF term `robustifies' the search for a stabilizing controller by ensuring that even highly sub-optimal polices will stabilize the system. We demonstrate our approach with two hardware examples where we learn stabilizing controllers for a cartpole and an A1 quadruped with only seconds and a few minutes of fine-tuning data, respectively. Furthermore, simulation benchmark studies show that obtaining stabilizing policies by optimizing our proposed costs requires orders of magnitude less data compared to standard cost designs.