NEARL: Non-Explicit Action Reinforcement Learning for Robotic Control

TL;DR

NEARL introduces a hierarchical reinforcement learning framework that avoids explicit actions, using a meta policy and inverse dynamics, enhancing safety and robustness in robotic control through imitation learning and adversarial training.

Contribution

The paper presents a novel hierarchical RL framework without explicit actions, integrating adversarial learning and inverse dynamics for safer, more effective robotic control.

Findings

Effective exploitation of state-only demonstrations for imitation learning

Enhanced stability and robustness in robotic control tasks

Successful application in simulation environments

Abstract

Traditionally, reinforcement learning methods predict the next action based on the current state. However, in many situations, directly applying actions to control systems or robots is dangerous and may lead to unexpected behaviors because action is rather low-level. In this paper, we propose a novel hierarchical reinforcement learning framework without explicit action. Our meta policy tries to manipulate the next optimal state and actual action is produced by the inverse dynamics model. To stabilize the training process, we integrate adversarial learning and information bottleneck into our framework. Under our framework, widely available state-only demonstrations can be exploited effectively for imitation learning. Also, prior knowledge and constraints can be applied to meta policy. We test our algorithm in simulation tasks and its combination with imitation learning. The experimental results show the reliability and robustness of our algorithms.