Combining Model-Based and Model-Free Updates for Trajectory-Centric Reinforcement Learning

TL;DR

This paper presents a novel approach that combines model-based and model-free reinforcement learning techniques, enabling efficient and effective policy learning for robotic manipulation tasks.

Contribution

The authors develop a unified framework integrating LQR-based model-based updates with path integral policy improvement, extendable to deep neural policies via guided policy search.

Findings

Achieves high sample efficiency comparable to model-based methods

Demonstrates superior performance on complex manipulation tasks

Validates effectiveness through both simulation and real-world experiments

Abstract

Reinforcement learning (RL) algorithms for real-world robotic applications need a data-efficient learning process and the ability to handle complex, unknown dynamical systems. These requirements are handled well by model-based and model-free RL approaches, respectively. In this work, we aim to combine the advantages of these two types of methods in a principled manner. By focusing on time-varying linear-Gaussian policies, we enable a model-based algorithm based on the linear quadratic regulator (LQR) that can be integrated into the model-free framework of path integral policy improvement (PI2). We can further combine our method with guided policy search (GPS) to train arbitrary parameterized policies such as deep neural networks. Our simulation and real-world experiments demonstrate that this method can solve challenging manipulation tasks with comparable or better performance than model-free methods while maintaining the sample efficiency of model-based methods. A video presenting our results is available at https://sites.google.com/site/icml17pilqr