Energy-Based Legged Robots Terrain Traversability Modeling via Deep Inverse Reinforcement Learning

This work reports ondeveloping a deep inverse reinforcement learning method for legged robots terrain traversability modeling that incorporates both exteroceptive and proprioceptive sensory data. Existing works use robot-agnostic exteroceptive environmental features or handcrafted kinematic features...

Full description

Saved in:
Bibliographic Details
Published in:IEEE robotics and automation letters 2022-10, Vol.7 (4), p.8807-8814
Main Authors: Gan, Lu, Grizzle, Jessy W., Eustice, Ryan M., Ghaffari, Maani
Format: Article
Language:English
Subjects:
Algorithms
Artificial neural networks
Behavioral sciences
Costs
Energy and environment-aware automation
Energy consumption
Energy management
Entropy
Inertial sensing devices
Kinematics
learning from demonstration
Legged locomotion
legged robots
Locomotion
Machine learning
Maximum entropy
Modelling
Robot sensing systems
Robots
Terrain
Trajectory
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This work reports ondeveloping a deep inverse reinforcement learning method for legged robots terrain traversability modeling that incorporates both exteroceptive and proprioceptive sensory data. Existing works use robot-agnostic exteroceptive environmental features or handcrafted kinematic features; instead, we propose to also learn robot-specific inertial features from proprioceptive sensory data for reward approximation in a single deep neural network. Incorporating the inertial features can improve the model fidelity and provide a reward that depends on the robot's state during deployment. We train the reward network using the Maximum Entropy Deep Inverse Reinforcement Learning (MEDIRL) algorithm and propose simultaneously minimizing a trajectory ranking loss to deal with the suboptimality of legged robot demonstrations. The demonstrated trajectories are ranked by locomotion energy consumption, in order to learn an energy-aware reward function and a more energy-efficient policy than demonstration. We evaluate our method using a dataset collected by an MIT Mini-Cheetah robot and a Mini-Cheetah simulator. The code is publicly available. 1
ISSN:2377-3766
2377-3766
DOI:10.1109/LRA.2022.3188100