Magnetic Field-Based Reward Shaping for Goal-Conditioned Reinforcement Learning

TL;DR

This paper introduces a magnetic field-inspired reward shaping method for goal-conditioned reinforcement learning, leveraging nonlinear magnetic field properties to improve learning efficiency in complex, dynamic environments.

Contribution

The paper proposes a novel magnetic field-based reward shaping technique that captures complex environmental information, enhancing sample efficiency in goal-conditioned RL tasks with dynamic targets and obstacles.

Findings

Outperforms existing reward shaping methods in simulated tasks.

Improves sample efficiency in real-world robotic manipulation.

Effectively handles dynamic environments with changing targets and obstacles.

Abstract

Goal-conditioned reinforcement learning (RL) is an interesting extension of the traditional RL framework, where the dynamic environment and reward sparsity can cause conventional learning algorithms to fail. Reward shaping is a practical approach to improving sample efficiency by embedding human domain knowledge into the learning process. Existing reward shaping methods for goal-conditioned RL are typically built on distance metrics with a linear and isotropic distribution, which may fail to provide sufficient information about the ever-changing environment with high complexity. This paper proposes a novel magnetic field-based reward shaping (MFRS) method for goal-conditioned RL tasks with dynamic target and obstacles. Inspired by the physical properties of magnets, we consider the target and obstacles as permanent magnets and establish the reward function according to the intensity values of the magnetic field generated by these magnets. The nonlinear and anisotropic distribution of the magnetic field intensity can provide more accessible and conducive information about the optimization landscape, thus introducing a more sophisticated magnetic reward compared to the distance-based setting. Further, we transform our magnetic reward to the form of potential-based reward shaping by learning a secondary potential function concurrently to ensure the optimal policy invariance of our method. Experiments results in both simulated and real-world robotic manipulation tasks demonstrate that MFRS outperforms relevant existing methods and effectively improves the sample efficiency of RL algorithms in goal-conditioned tasks with various dynamics of the target and obstacles.