Hyperdimensional computing in industrial systems: the use-case of distributed fault isolation in a power plant

Denis Kleyko, Evgeny Osipov, Nikolaos Papakonstantinou, Valeriy Vyatkin

    Research output: Contribution to journalArticleScientificpeer-review

    3 Citations (Scopus)

    Abstract

    This article presents an approach for distributed fault isolation in a generic system of systems. The proposed approach is based on the principles of hyperdimensional computing. In particular, the recently proposed method called Holographic Graph Neuron is used. We present a distributed version of Holographic Graph Neuron and evaluate its performance on the problem of fault isolation in a complex power plant model. Compared to conventional machine learning methods applied in the context of the same scenario the proposed approach shows comparable performance while being distributed and requiring simple binary operations, which allow for a fast and efficient implementation in a hardware.

    Original languageEnglish
    Pages (from-to)30766-30777
    Number of pages12
    JournalIEEE Access
    Volume6
    DOIs
    Publication statusPublished - 2018
    MoE publication typeA1 Journal article-refereed

    Fingerprint

    Neurons
    Power plants
    Learning systems
    Hardware
    System of systems

    Keywords

    • Automation
    • complex systems
    • Computational modeling
    • distributed fault isolation
    • distributed representation
    • Feature extraction
    • Holographic Graph Neuron
    • hyperdimensional computing
    • Machine learning
    • machine learning
    • Neurons
    • Sensors
    • Training
    • Vector Symbolic Architectures

    Cite this

    @article{7727e691d7a34b23a8887bb10b032efc,
    title = "Hyperdimensional computing in industrial systems: the use-case of distributed fault isolation in a power plant",
    abstract = "This article presents an approach for distributed fault isolation in a generic system of systems. The proposed approach is based on the principles of hyperdimensional computing. In particular, the recently proposed method called Holographic Graph Neuron is used. We present a distributed version of Holographic Graph Neuron and evaluate its performance on the problem of fault isolation in a complex power plant model. Compared to conventional machine learning methods applied in the context of the same scenario the proposed approach shows comparable performance while being distributed and requiring simple binary operations, which allow for a fast and efficient implementation in a hardware.",
    keywords = "Automation, complex systems, Computational modeling, distributed fault isolation, distributed representation, Feature extraction, Holographic Graph Neuron, hyperdimensional computing, Machine learning, machine learning, Neurons, Sensors, Training, Vector Symbolic Architectures",
    author = "Denis Kleyko and Evgeny Osipov and Nikolaos Papakonstantinou and Valeriy Vyatkin",
    year = "2018",
    doi = "10.1109/ACCESS.2018.2840128",
    language = "English",
    volume = "6",
    pages = "30766--30777",
    journal = "IEEE Access",
    issn = "2169-3536",
    publisher = "IEEE Institute of Electrical and Electronic Engineers",

    }

    Hyperdimensional computing in industrial systems : the use-case of distributed fault isolation in a power plant. / Kleyko, Denis; Osipov, Evgeny; Papakonstantinou, Nikolaos; Vyatkin, Valeriy.

    In: IEEE Access, Vol. 6, 2018, p. 30766-30777.

    Research output: Contribution to journalArticleScientificpeer-review

    TY - JOUR

    T1 - Hyperdimensional computing in industrial systems

    T2 - the use-case of distributed fault isolation in a power plant

    AU - Kleyko, Denis

    AU - Osipov, Evgeny

    AU - Papakonstantinou, Nikolaos

    AU - Vyatkin, Valeriy

    PY - 2018

    Y1 - 2018

    N2 - This article presents an approach for distributed fault isolation in a generic system of systems. The proposed approach is based on the principles of hyperdimensional computing. In particular, the recently proposed method called Holographic Graph Neuron is used. We present a distributed version of Holographic Graph Neuron and evaluate its performance on the problem of fault isolation in a complex power plant model. Compared to conventional machine learning methods applied in the context of the same scenario the proposed approach shows comparable performance while being distributed and requiring simple binary operations, which allow for a fast and efficient implementation in a hardware.

    AB - This article presents an approach for distributed fault isolation in a generic system of systems. The proposed approach is based on the principles of hyperdimensional computing. In particular, the recently proposed method called Holographic Graph Neuron is used. We present a distributed version of Holographic Graph Neuron and evaluate its performance on the problem of fault isolation in a complex power plant model. Compared to conventional machine learning methods applied in the context of the same scenario the proposed approach shows comparable performance while being distributed and requiring simple binary operations, which allow for a fast and efficient implementation in a hardware.

    KW - Automation

    KW - complex systems

    KW - Computational modeling

    KW - distributed fault isolation

    KW - distributed representation

    KW - Feature extraction

    KW - Holographic Graph Neuron

    KW - hyperdimensional computing

    KW - Machine learning

    KW - machine learning

    KW - Neurons

    KW - Sensors

    KW - Training

    KW - Vector Symbolic Architectures

    UR - http://www.scopus.com/inward/record.url?scp=85047613488&partnerID=8YFLogxK

    U2 - 10.1109/ACCESS.2018.2840128

    DO - 10.1109/ACCESS.2018.2840128

    M3 - Article

    AN - SCOPUS:85047613488

    VL - 6

    SP - 30766

    EP - 30777

    JO - IEEE Access

    JF - IEEE Access

    SN - 2169-3536

    ER -