Dynamic positioning (DP) is one of the key technologies in the various offshore operations. This paper investigates machine learning (ML) techniques, more specifically reinforcement learning (RL) in DP applications for deepwater drilling rigs. An RL-based DP control algorithm with the deep Q-learning network is proposed. To guarantee the safe operation of the underwater riser, a comprehensive reward function with respect to the top and bottom angles of the riser is designed. Compared to traditional control methods, the proposed RL-based algorithm shows good positioning performance while maintaining the operating limits of the underwater riser.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Deep reinforcement learning in dynamic positioning control: by rewarding small response of riser angles


    Contributors:
    Wang, Fang (author) / Bai, Yong (author) / Bai, Jie (author) / Zhao, Liang (author)

    Published in:

    Ships and Offshore Structures ; 18 , 11 ; 1497-1504


    Publication date :

    2023-11-02


    Size :

    8 pages




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    Unknown




    Modelling and Control of Riser Angles and Stresses in Dynamic Positioning

    Sorensen, A. J. / Leira, B. / Strand, J. P. et al. | British Library Conference Proceedings | 2001


    Dynamic Positioning using Deep Reinforcement Learning

    Øvereng, Simen Sem / Nguyen, Dong Trong | BASE | 2021

    Free access

    Control of marine riser end angles by position mooring

    Nguyen, Dat H. / Nguyen, Dong T. / Quek, Ser T. et al. | Tema Archive | 2010


    Deep Reinforcement Learning for Dynamic Grasping

    Ström, Andreas | BASE | 2022

    Free access

    Rewarding Environmental Excellence

    Finch, Ginny | Online Contents | 1995