Efficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTM

Abstract To address the challenges of sample utilization efficiency and managing temporal dependencies, this paper proposes an efficient path planning method for mobile robot in dynamic environments based on an improved twin delayed deep deterministic policy gradient (TD3) algorithm. The proposed me...

Full description

Saved in:
Bibliographic Details
Main Authors: Yunhan Lin, Zhijie Zhang, Yijian Tan, Hao Fu, Huasong Min
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-02244-z
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849688085250441216
author Yunhan Lin
Zhijie Zhang
Yijian Tan
Hao Fu
Huasong Min
author_facet Yunhan Lin
Zhijie Zhang
Yijian Tan
Hao Fu
Huasong Min
author_sort Yunhan Lin
collection DOAJ
description Abstract To address the challenges of sample utilization efficiency and managing temporal dependencies, this paper proposes an efficient path planning method for mobile robot in dynamic environments based on an improved twin delayed deep deterministic policy gradient (TD3) algorithm. The proposed method, named PL-TD3, integrates prioritized experience replay (PER) and long short-term memory (LSTM) neural networks, which enhance both sample efficiency and the ability to handle time-series data. To verify the effectiveness of the proposed method, simulation and practical experiments were designed and conducted. In the simulation experiments, both static and dynamic obstacles were included in the test environment, along with experiments to assess generalization capabilities. The algorithm demonstrated superior performance in terms of both execution time and path efficiency. The practical experiments, based on the assumptions from the simulation tests, further confirmed that PL-TD3 has improved the effectiveness and robustness of path planning for mobile robot in dynamic environments.
format Article
id doaj-art-70a6647a072f4f80bada8f87da58ed87
institution DOAJ
issn 2045-2322
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-70a6647a072f4f80bada8f87da58ed872025-08-20T03:22:08ZengNature PortfolioScientific Reports2045-23222025-05-0115111610.1038/s41598-025-02244-zEfficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTMYunhan Lin0Zhijie Zhang1Yijian Tan2Hao Fu3Huasong Min4School of Computer Science and Technology, Wuhan University of Science and TechnologySchool of Computer Science and Technology, Wuhan University of Science and TechnologySchool of Computer Science and Technology, Wuhan University of Science and TechnologySchool of Computer Science and Technology, Wuhan University of Science and TechnologyInstitute of Robotics and Intelligent Systems, Wuhan University of Science and TechnologyAbstract To address the challenges of sample utilization efficiency and managing temporal dependencies, this paper proposes an efficient path planning method for mobile robot in dynamic environments based on an improved twin delayed deep deterministic policy gradient (TD3) algorithm. The proposed method, named PL-TD3, integrates prioritized experience replay (PER) and long short-term memory (LSTM) neural networks, which enhance both sample efficiency and the ability to handle time-series data. To verify the effectiveness of the proposed method, simulation and practical experiments were designed and conducted. In the simulation experiments, both static and dynamic obstacles were included in the test environment, along with experiments to assess generalization capabilities. The algorithm demonstrated superior performance in terms of both execution time and path efficiency. The practical experiments, based on the assumptions from the simulation tests, further confirmed that PL-TD3 has improved the effectiveness and robustness of path planning for mobile robot in dynamic environments.https://doi.org/10.1038/s41598-025-02244-zPath planning in dynamic environmentTwin delayed deep deterministic policy gradient (TD3) algorithmReinforcement learningPrioritized experience replay (PER)Long short-term memory (LSTM )
spellingShingle Yunhan Lin
Zhijie Zhang
Yijian Tan
Hao Fu
Huasong Min
Efficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTM
Scientific Reports
Path planning in dynamic environment
Twin delayed deep deterministic policy gradient (TD3) algorithm
Reinforcement learning
Prioritized experience replay (PER)
Long short-term memory (LSTM )
title Efficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTM
title_full Efficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTM
title_fullStr Efficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTM
title_full_unstemmed Efficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTM
title_short Efficient TD3 based path planning of mobile robot in dynamic environments using prioritized experience replay and LSTM
title_sort efficient td3 based path planning of mobile robot in dynamic environments using prioritized experience replay and lstm
topic Path planning in dynamic environment
Twin delayed deep deterministic policy gradient (TD3) algorithm
Reinforcement learning
Prioritized experience replay (PER)
Long short-term memory (LSTM )
url https://doi.org/10.1038/s41598-025-02244-z
work_keys_str_mv AT yunhanlin efficienttd3basedpathplanningofmobilerobotindynamicenvironmentsusingprioritizedexperiencereplayandlstm
AT zhijiezhang efficienttd3basedpathplanningofmobilerobotindynamicenvironmentsusingprioritizedexperiencereplayandlstm
AT yijiantan efficienttd3basedpathplanningofmobilerobotindynamicenvironmentsusingprioritizedexperiencereplayandlstm
AT haofu efficienttd3basedpathplanningofmobilerobotindynamicenvironmentsusingprioritizedexperiencereplayandlstm
AT huasongmin efficienttd3basedpathplanningofmobilerobotindynamicenvironmentsusingprioritizedexperiencereplayandlstm