Comparative study of different artificial neural networks methodoiogies on static solar photovoltaic Module

Xu HUANG, Dat TRAN

    Research output: Contribution to journalArticle

    3 Citations (Scopus)

    Abstract

    The aim of this paper is to present a comparative study of the different artificial neural networks modeling approach on the static solar photovoltaic (SSPV) model under variable input parameters and conditions. Solar energy generation has widely based its power generation on static solar photovoltaic system generation due to its non-involvement in recurrent maintenance costs and non-replacement of tracking devices after initial installation. The artificial neural networks have proven to be a reliable software application tool capable of providing solutions to complex non-linear mathematical equations and complicated models. It provides accurate prediction and generalisation of the system at high speed. The results of the comparative artificial neural networks indicate the performance of the different methodologies and their respective output efficiency characteristics for evaluation. The autocorrelation coefficient obtained from each methodology gives an accuracy of approximately 99% with negligible mean square error (MSE) for the output variables. The performance of these methodologies are compared and presented on a very large scale for all operating conditions to confirm the validity of the model.
    Original languageEnglish
    Pages (from-to)674-685
    Number of pages12
    JournalInternational Journal of Emerging Technology and Advanced Engineering
    Volume4
    Issue number10
    Publication statusPublished - 2014

    Fingerprint

    Neural networks
    Autocorrelation
    Application programs
    Mean square error
    Solar energy
    Power generation
    Costs

    Cite this

    @article{eb9cb242a2d34f1aaee4f4f449c05607,
    title = "Comparative study of different artificial neural networks methodoiogies on static solar photovoltaic Module",
    abstract = "The aim of this paper is to present a comparative study of the different artificial neural networks modeling approach on the static solar photovoltaic (SSPV) model under variable input parameters and conditions. Solar energy generation has widely based its power generation on static solar photovoltaic system generation due to its non-involvement in recurrent maintenance costs and non-replacement of tracking devices after initial installation. The artificial neural networks have proven to be a reliable software application tool capable of providing solutions to complex non-linear mathematical equations and complicated models. It provides accurate prediction and generalisation of the system at high speed. The results of the comparative artificial neural networks indicate the performance of the different methodologies and their respective output efficiency characteristics for evaluation. The autocorrelation coefficient obtained from each methodology gives an accuracy of approximately 99{\%} with negligible mean square error (MSE) for the output variables. The performance of these methodologies are compared and presented on a very large scale for all operating conditions to confirm the validity of the model.",
    keywords = "Artificial-neural-network, renewable-energy, maximum-power-point",
    author = "Xu HUANG and Dat TRAN",
    year = "2014",
    language = "English",
    volume = "4",
    pages = "674--685",
    journal = "International Journal of Emerging Technology and Advanced Engineering",
    issn = "2250-2459",
    number = "10",

    }

    TY - JOUR

    T1 - Comparative study of different artificial neural networks methodoiogies on static solar photovoltaic Module

    AU - HUANG, Xu

    AU - TRAN, Dat

    PY - 2014

    Y1 - 2014

    N2 - The aim of this paper is to present a comparative study of the different artificial neural networks modeling approach on the static solar photovoltaic (SSPV) model under variable input parameters and conditions. Solar energy generation has widely based its power generation on static solar photovoltaic system generation due to its non-involvement in recurrent maintenance costs and non-replacement of tracking devices after initial installation. The artificial neural networks have proven to be a reliable software application tool capable of providing solutions to complex non-linear mathematical equations and complicated models. It provides accurate prediction and generalisation of the system at high speed. The results of the comparative artificial neural networks indicate the performance of the different methodologies and their respective output efficiency characteristics for evaluation. The autocorrelation coefficient obtained from each methodology gives an accuracy of approximately 99% with negligible mean square error (MSE) for the output variables. The performance of these methodologies are compared and presented on a very large scale for all operating conditions to confirm the validity of the model.

    AB - The aim of this paper is to present a comparative study of the different artificial neural networks modeling approach on the static solar photovoltaic (SSPV) model under variable input parameters and conditions. Solar energy generation has widely based its power generation on static solar photovoltaic system generation due to its non-involvement in recurrent maintenance costs and non-replacement of tracking devices after initial installation. The artificial neural networks have proven to be a reliable software application tool capable of providing solutions to complex non-linear mathematical equations and complicated models. It provides accurate prediction and generalisation of the system at high speed. The results of the comparative artificial neural networks indicate the performance of the different methodologies and their respective output efficiency characteristics for evaluation. The autocorrelation coefficient obtained from each methodology gives an accuracy of approximately 99% with negligible mean square error (MSE) for the output variables. The performance of these methodologies are compared and presented on a very large scale for all operating conditions to confirm the validity of the model.

    KW - Artificial-neural-network

    KW - renewable-energy

    KW - maximum-power-point

    M3 - Article

    VL - 4

    SP - 674

    EP - 685

    JO - International Journal of Emerging Technology and Advanced Engineering

    JF - International Journal of Emerging Technology and Advanced Engineering

    SN - 2250-2459

    IS - 10

    ER -