Accelerating Model-Based Reinforcement Learning using Non-Linear Trajectory Optimization

TL;DR

This paper introduces EB-MC-PILCO, a method that combines MC-PILCO with iLQR to accelerate policy optimization in model-based reinforcement learning, demonstrated on the cart-pole task.

Contribution

It integrates iLQR into MC-PILCO to improve convergence speed and exploration efficiency in nonlinear control tasks.

Findings

Up to 45.9% reduction in execution time.

Maintains 100% success rate across trials.

Faster convergence even with fewer iterations.

Abstract

This paper addresses the slow policy optimization convergence of Monte Carlo Probabilistic Inference for Learning Control (MC-PILCO), a state-of-the-art model-based reinforcement learning (MBRL) algorithm, by integrating it with iterative Linear Quadratic Regulator (iLQR), a fast trajectory optimization method suitable for nonlinear systems. The proposed method, Exploration-Boosted MC-PILCO (EB-MC-PILCO), leverages iLQR to generate informative, exploratory trajectories and initialize the policy, significantly reducing the number of required optimization steps. Experiments on the cart-pole task demonstrate that EB-MC-PILCO accelerates convergence compared to standard MC-PILCO, achieving up to $\bm{45.9\%}$ reduction in execution time when both methods solve the task in four trials. EB-MC-PILCO also maintains a $\bm{100\%}$ success rate across trials while solving the task faster, even in cases where MC-PILCO converges in fewer iterations.