Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov models

Abstract With the global energy structure shifting towards clean and efficient hydrogen energy, the safety management issues of hydrogen refueling stations are becoming increasingly prominent. To address these issues, a hydrogen leak localization algorithm for hydrogen refueling stations based on a...

Full description

Saved in:
Bibliographic Details
Main Authors: Jun Wang, Lei Wang, Ding Wang, Xiang Yu
Format: Article
Language:English
Published: SpringerOpen 2025-07-01
Series:Sustainable Energy Research
Subjects:
Online Access:https://doi.org/10.1186/s40807-025-00186-8
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849342658231664640
author Jun Wang
Lei Wang
Ding Wang
Xiang Yu
author_facet Jun Wang
Lei Wang
Ding Wang
Xiang Yu
author_sort Jun Wang
collection DOAJ
description Abstract With the global energy structure shifting towards clean and efficient hydrogen energy, the safety management issues of hydrogen refueling stations are becoming increasingly prominent. To address these issues, a hydrogen leak localization algorithm for hydrogen refueling stations based on a combination of reinforcement learning and hidden Markov models is proposed. This method combines hidden Markov model to construct a probability distribution model for hydrogen leakage and diffusion, simulates the propagation probability of hydrogen in different grid cells, and uses reinforcement learning to achieve fast and accurate localization of hydrogen leakage events. The outcomes denoted that the training accuracy reached 95.2%, with an F1 value of 0.961, indicating its high accuracy in hydrogen leak localization. When the wind speed was 0.8 m/s, the mean square error of the raised method was 0.03, and when the wind speed was 1.0 m/s, the mean square error of the raised method was 0.04, proving its good robustness. After 50 localization experiments, the proposed algorithm achieves a localization success rate of 93.7% and an average computation time of 42.8 s, further demonstrating its high accuracy and computational efficiency. The proposed hydrogen leakage location algorithm has improved the accuracy and efficiency of hydrogen leakage location, providing scientific basis and technical guarantee for the safe operation of future hydrogen refueling stations.
format Article
id doaj-art-14e4fffb413e46daba40f91aae3084bc
institution Kabale University
issn 2731-9237
language English
publishDate 2025-07-01
publisher SpringerOpen
record_format Article
series Sustainable Energy Research
spelling doaj-art-14e4fffb413e46daba40f91aae3084bc2025-08-20T03:43:20ZengSpringerOpenSustainable Energy Research2731-92372025-07-0112111210.1186/s40807-025-00186-8Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov modelsJun Wang0Lei Wang1Ding Wang2Xiang Yu3College of Electronics and Information Engineering, Tongji UniversityCollege of Electronics and Information Engineering, Tongji UniversityCollege of Electronics and Information Engineering, Tongji UniversityCollege of Electronics and Information Engineering, Tongji UniversityAbstract With the global energy structure shifting towards clean and efficient hydrogen energy, the safety management issues of hydrogen refueling stations are becoming increasingly prominent. To address these issues, a hydrogen leak localization algorithm for hydrogen refueling stations based on a combination of reinforcement learning and hidden Markov models is proposed. This method combines hidden Markov model to construct a probability distribution model for hydrogen leakage and diffusion, simulates the propagation probability of hydrogen in different grid cells, and uses reinforcement learning to achieve fast and accurate localization of hydrogen leakage events. The outcomes denoted that the training accuracy reached 95.2%, with an F1 value of 0.961, indicating its high accuracy in hydrogen leak localization. When the wind speed was 0.8 m/s, the mean square error of the raised method was 0.03, and when the wind speed was 1.0 m/s, the mean square error of the raised method was 0.04, proving its good robustness. After 50 localization experiments, the proposed algorithm achieves a localization success rate of 93.7% and an average computation time of 42.8 s, further demonstrating its high accuracy and computational efficiency. The proposed hydrogen leakage location algorithm has improved the accuracy and efficiency of hydrogen leakage location, providing scientific basis and technical guarantee for the safe operation of future hydrogen refueling stations.https://doi.org/10.1186/s40807-025-00186-8Reinforcement learningHidden MarkovHydrogen refueling stationHydrogen gas leakageLocation
spellingShingle Jun Wang
Lei Wang
Ding Wang
Xiang Yu
Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov models
Sustainable Energy Research
Reinforcement learning
Hidden Markov
Hydrogen refueling station
Hydrogen gas leakage
Location
title Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov models
title_full Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov models
title_fullStr Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov models
title_full_unstemmed Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov models
title_short Hydrogen leakage localization technology in hydrogen refueling stations combining RL and hidden Markov models
title_sort hydrogen leakage localization technology in hydrogen refueling stations combining rl and hidden markov models
topic Reinforcement learning
Hidden Markov
Hydrogen refueling station
Hydrogen gas leakage
Location
url https://doi.org/10.1186/s40807-025-00186-8
work_keys_str_mv AT junwang hydrogenleakagelocalizationtechnologyinhydrogenrefuelingstationscombiningrlandhiddenmarkovmodels
AT leiwang hydrogenleakagelocalizationtechnologyinhydrogenrefuelingstationscombiningrlandhiddenmarkovmodels
AT dingwang hydrogenleakagelocalizationtechnologyinhydrogenrefuelingstationscombiningrlandhiddenmarkovmodels
AT xiangyu hydrogenleakagelocalizationtechnologyinhydrogenrefuelingstationscombiningrlandhiddenmarkovmodels