Hybrid DQN-TD3 Reinforcement Learning for Autonomous Navigation in Dynamic Environments

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: He, Xiaoyi, Chen, Danggui, Zhang, Zhenshuo, Bai, Zimeng
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908620708380672
author He, Xiaoyi
Chen, Danggui
Zhang, Zhenshuo
Bai, Zimeng
author_facet He, Xiaoyi
Chen, Danggui
Zhang, Zhenshuo
Bai, Zimeng
contents This paper presents a hierarchical path-planning and control framework that combines a high-level Deep Q-Network (DQN) for discrete sub-goal selection with a low-level Twin Delayed Deep Deterministic Policy Gradient (TD3) controller for continuous actuation. The high-level module selects behaviors and sub-goals; the low-level module executes smooth velocity commands. We design a practical reward shaping scheme (direction, distance, obstacle avoidance, action smoothness, collision penalty, time penalty, and progress), together with a LiDAR-based safety gate that prevents unsafe motions. The system is implemented in ROS + Gazebo (TurtleBot3) and evaluated with PathBench metrics, including success rate, collision rate, path efficiency, and re-planning efficiency, in dynamic and partially observable environments. Experiments show improved success rate and sample efficiency over single-algorithm baselines (DQN or TD3 alone) and rule-based planners, with better generalization to unseen obstacle configurations and reduced abrupt control changes. Code and evaluation scripts are available at the project repository.
format Preprint
id arxiv_https___arxiv_org_abs_2510_26646
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Hybrid DQN-TD3 Reinforcement Learning for Autonomous Navigation in Dynamic Environments
He, Xiaoyi
Chen, Danggui
Zhang, Zhenshuo
Bai, Zimeng
Robotics
Artificial Intelligence
Machine Learning
This paper presents a hierarchical path-planning and control framework that combines a high-level Deep Q-Network (DQN) for discrete sub-goal selection with a low-level Twin Delayed Deep Deterministic Policy Gradient (TD3) controller for continuous actuation. The high-level module selects behaviors and sub-goals; the low-level module executes smooth velocity commands. We design a practical reward shaping scheme (direction, distance, obstacle avoidance, action smoothness, collision penalty, time penalty, and progress), together with a LiDAR-based safety gate that prevents unsafe motions. The system is implemented in ROS + Gazebo (TurtleBot3) and evaluated with PathBench metrics, including success rate, collision rate, path efficiency, and re-planning efficiency, in dynamic and partially observable environments. Experiments show improved success rate and sample efficiency over single-algorithm baselines (DQN or TD3 alone) and rule-based planners, with better generalization to unseen obstacle configurations and reduced abrupt control changes. Code and evaluation scripts are available at the project repository.
title Hybrid DQN-TD3 Reinforcement Learning for Autonomous Navigation in Dynamic Environments
topic Robotics
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2510.26646