Imitation of Demonstrations Using Bayesian Filtering with Nonparametric Data-Driven Models

Nurali Virani, Devesh K. Jha, Zhenyuan Yuan, Ishana Shekhawat, Asok Ray

    Research output: Contribution to journalArticle

    1 Citation (Scopus)

    Abstract

    This paper addresses the problem of learning dynamic models of hybrid systems from demonstrations and then the problem of imitation of those demonstrations by using Bayesian filtering. A linear programming-based approach is used to develop nonparametric kernel-based conditional density estimation technique to infer accurate and concise dynamic models of system evolution from data. The training data for these models have been acquired from demonstrations by teleoperation. The trained data-driven models for mode-dependent state evolution and state-dependent mode evolution are then used online for imitation of demonstrated tasks via particle filtering. The results of simulation and experimental validation with a hexapod robot are reported to establish generalization of the proposed learning and control algorithms.

    Original languageEnglish (US)
    Article number030906
    JournalJournal of Dynamic Systems, Measurement and Control, Transactions of the ASME
    Volume140
    Issue number3
    DOIs
    StatePublished - Mar 1 2018

    Fingerprint

    Demonstrations
    Dynamic models
    dynamic models
    learning
    Remote control
    Hybrid systems
    linear programming
    Linear programming
    Learning algorithms
    robots
    Robots
    education
    simulation

    All Science Journal Classification (ASJC) codes

    • Control and Systems Engineering
    • Information Systems
    • Instrumentation
    • Mechanical Engineering
    • Computer Science Applications

    Cite this

    @article{09eae402692e417abd949f7895945365,
    title = "Imitation of Demonstrations Using Bayesian Filtering with Nonparametric Data-Driven Models",
    abstract = "This paper addresses the problem of learning dynamic models of hybrid systems from demonstrations and then the problem of imitation of those demonstrations by using Bayesian filtering. A linear programming-based approach is used to develop nonparametric kernel-based conditional density estimation technique to infer accurate and concise dynamic models of system evolution from data. The training data for these models have been acquired from demonstrations by teleoperation. The trained data-driven models for mode-dependent state evolution and state-dependent mode evolution are then used online for imitation of demonstrated tasks via particle filtering. The results of simulation and experimental validation with a hexapod robot are reported to establish generalization of the proposed learning and control algorithms.",
    author = "Nurali Virani and Jha, {Devesh K.} and Zhenyuan Yuan and Ishana Shekhawat and Asok Ray",
    year = "2018",
    month = "3",
    day = "1",
    doi = "10.1115/1.4037782",
    language = "English (US)",
    volume = "140",
    journal = "Journal of Dynamic Systems, Measurement and Control, Transactions of the ASME",
    issn = "0022-0434",
    publisher = "American Society of Mechanical Engineers(ASME)",
    number = "3",

    }

    Imitation of Demonstrations Using Bayesian Filtering with Nonparametric Data-Driven Models. / Virani, Nurali; Jha, Devesh K.; Yuan, Zhenyuan; Shekhawat, Ishana; Ray, Asok.

    In: Journal of Dynamic Systems, Measurement and Control, Transactions of the ASME, Vol. 140, No. 3, 030906, 01.03.2018.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Imitation of Demonstrations Using Bayesian Filtering with Nonparametric Data-Driven Models

    AU - Virani, Nurali

    AU - Jha, Devesh K.

    AU - Yuan, Zhenyuan

    AU - Shekhawat, Ishana

    AU - Ray, Asok

    PY - 2018/3/1

    Y1 - 2018/3/1

    N2 - This paper addresses the problem of learning dynamic models of hybrid systems from demonstrations and then the problem of imitation of those demonstrations by using Bayesian filtering. A linear programming-based approach is used to develop nonparametric kernel-based conditional density estimation technique to infer accurate and concise dynamic models of system evolution from data. The training data for these models have been acquired from demonstrations by teleoperation. The trained data-driven models for mode-dependent state evolution and state-dependent mode evolution are then used online for imitation of demonstrated tasks via particle filtering. The results of simulation and experimental validation with a hexapod robot are reported to establish generalization of the proposed learning and control algorithms.

    AB - This paper addresses the problem of learning dynamic models of hybrid systems from demonstrations and then the problem of imitation of those demonstrations by using Bayesian filtering. A linear programming-based approach is used to develop nonparametric kernel-based conditional density estimation technique to infer accurate and concise dynamic models of system evolution from data. The training data for these models have been acquired from demonstrations by teleoperation. The trained data-driven models for mode-dependent state evolution and state-dependent mode evolution are then used online for imitation of demonstrated tasks via particle filtering. The results of simulation and experimental validation with a hexapod robot are reported to establish generalization of the proposed learning and control algorithms.

    UR - http://www.scopus.com/inward/record.url?scp=85034033941&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85034033941&partnerID=8YFLogxK

    U2 - 10.1115/1.4037782

    DO - 10.1115/1.4037782

    M3 - Article

    VL - 140

    JO - Journal of Dynamic Systems, Measurement and Control, Transactions of the ASME

    JF - Journal of Dynamic Systems, Measurement and Control, Transactions of the ASME

    SN - 0022-0434

    IS - 3

    M1 - 030906

    ER -