LyEvO: Lyapunov-Guided Evolutionary Optimization for Safe and Robust Sim-to-Real Policy Learning

2026-08-06

AI Summary

This work addresses the challenges of synthesizing safe and robust controllers through simulation-based training and evaluating their readiness for real-world deployment. It proposes a novel integrated approach that synergistically combines constrained evolutionary optimization, statistical model checking (SMC), and Lyapunov stability analysis. For the first time, these three components are jointly leveraged to iteratively refine control policies using prior knowledge of system dynamics and systematically expand regions of guaranteed stability, thereby establishing a quantifiable criterion for deployment readiness. Experimental results on both Cartpole and 3D quadrotor platforms demonstrate that the proposed method effectively enables safe, robust policy transfer from simulation to reality.

Abstract

Training controllers that are safe and robust in simulation, and systematically assessing their readiness for real-world deployment, remain key challenges in sim-to-real transfer. To address this, we propose LyEvO, a physics-grounded framework that combines constrained Evolutionary Optimization and Statistical Model Checking (SMC)-based verification with Lyapunov-based stability analysis. Leveraging prior knowledge of the system dynamics, LyEvO uses Lyapunov analysis to compute an initial candidate stability region. An iterative loop then uses operational scenarios drawn from this region to jointly optimize and statistically verify a policy, and subsequently expands the region's boundaries based on the verification outcome. This integrated procedure provides a practical criterion for assessing deployment readiness. We evaluate LyEvO on Cartpole and 3D Quadrotor benchmarks through extensive simulations and targeted real-world experiments, demonstrating safe and robust sim-to-real transfer.

Problem

• sim-to-real transfer
• safe control
• robust policy
• deployment readiness
• Lyapunov stability

Innovation

• Lyapunov stability
• Evolutionary Optimization
• Statistical Model Checking
• Sim-to-Real Transfer
• Safety Verification