Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities

B. Kermanshahi, Y. Akiyama, R. Yokoyama, M. Asari, K. Takahashi

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    In this paper, a recurrent neural network (RNN) is applied for long-term load forecasting. The proposed RNN is trained with the past 20 years (1975-1994) of actual data and is tested for target years (1995-1997, 2000, and 2005). In addition to the target year load forecasting, a sliding window training method is proposed for continuous retraining of the RNN. The actual data of 9 Japanese power utilities is used for forecasting the loads of 1975 to 1994. However, forecasted data is applied for forecasting the loads beyond 1994. Since the weather condition data is not available for longer than two weeks ahead, a sensitivity program is developed to produce the future temperature from the present one. Very reasonable results have been obtained for the seen (inner sample) and unseen (out-of-sample or loads of target years) data. In this study, total system load forecast reflecting current and future trends, tempered with good judgment which is the key to all planning, indeed financial success is carried out for 9 power utilities in Japan. The obtained results of this study will be useful for other country's utilities.

    Original languageEnglish
    Title of host publicationProceedings of the Universities Power Engineering Conference
    Place of PublicationIraklio, Greece
    PublisherTechnological Educational Institute
    Pages895-898
    Number of pages4
    Volume2
    Publication statusPublished - 1997
    EventProceedings of the 1997 32nd Univertsities Power Engineering Conference, UPEC'97. Part 2 (of 2) - Manchester, UK
    Duration: 1997 Sep 101997 Sep 12

    Other

    OtherProceedings of the 1997 32nd Univertsities Power Engineering Conference, UPEC'97. Part 2 (of 2)
    CityManchester, UK
    Period97/9/1097/9/12

    Fingerprint

    Recurrent neural networks
    Planning
    Temperature

    ASJC Scopus subject areas

    • Energy(all)
    • Engineering(all)

    Cite this

    Kermanshahi, B., Akiyama, Y., Yokoyama, R., Asari, M., & Takahashi, K. (1997). Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities. In Proceedings of the Universities Power Engineering Conference (Vol. 2, pp. 895-898). Iraklio, Greece: Technological Educational Institute.

    Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities. / Kermanshahi, B.; Akiyama, Y.; Yokoyama, R.; Asari, M.; Takahashi, K.

    Proceedings of the Universities Power Engineering Conference. Vol. 2 Iraklio, Greece : Technological Educational Institute, 1997. p. 895-898.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Kermanshahi, B, Akiyama, Y, Yokoyama, R, Asari, M & Takahashi, K 1997, Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities. in Proceedings of the Universities Power Engineering Conference. vol. 2, Technological Educational Institute, Iraklio, Greece, pp. 895-898, Proceedings of the 1997 32nd Univertsities Power Engineering Conference, UPEC'97. Part 2 (of 2), Manchester, UK, 97/9/10.
    Kermanshahi B, Akiyama Y, Yokoyama R, Asari M, Takahashi K. Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities. In Proceedings of the Universities Power Engineering Conference. Vol. 2. Iraklio, Greece: Technological Educational Institute. 1997. p. 895-898
    Kermanshahi, B. ; Akiyama, Y. ; Yokoyama, R. ; Asari, M. ; Takahashi, K. / Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities. Proceedings of the Universities Power Engineering Conference. Vol. 2 Iraklio, Greece : Technological Educational Institute, 1997. pp. 895-898
    @inproceedings{e499cfd8f7cf4ce583961c00099fa512,
    title = "Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities",
    abstract = "In this paper, a recurrent neural network (RNN) is applied for long-term load forecasting. The proposed RNN is trained with the past 20 years (1975-1994) of actual data and is tested for target years (1995-1997, 2000, and 2005). In addition to the target year load forecasting, a sliding window training method is proposed for continuous retraining of the RNN. The actual data of 9 Japanese power utilities is used for forecasting the loads of 1975 to 1994. However, forecasted data is applied for forecasting the loads beyond 1994. Since the weather condition data is not available for longer than two weeks ahead, a sensitivity program is developed to produce the future temperature from the present one. Very reasonable results have been obtained for the seen (inner sample) and unseen (out-of-sample or loads of target years) data. In this study, total system load forecast reflecting current and future trends, tempered with good judgment which is the key to all planning, indeed financial success is carried out for 9 power utilities in Japan. The obtained results of this study will be useful for other country's utilities.",
    author = "B. Kermanshahi and Y. Akiyama and R. Yokoyama and M. Asari and K. Takahashi",
    year = "1997",
    language = "English",
    volume = "2",
    pages = "895--898",
    booktitle = "Proceedings of the Universities Power Engineering Conference",
    publisher = "Technological Educational Institute",

    }

    TY - GEN

    T1 - Recurrent neural network for forecasting next 10 years loads of 9 Japanese utilities

    AU - Kermanshahi, B.

    AU - Akiyama, Y.

    AU - Yokoyama, R.

    AU - Asari, M.

    AU - Takahashi, K.

    PY - 1997

    Y1 - 1997

    N2 - In this paper, a recurrent neural network (RNN) is applied for long-term load forecasting. The proposed RNN is trained with the past 20 years (1975-1994) of actual data and is tested for target years (1995-1997, 2000, and 2005). In addition to the target year load forecasting, a sliding window training method is proposed for continuous retraining of the RNN. The actual data of 9 Japanese power utilities is used for forecasting the loads of 1975 to 1994. However, forecasted data is applied for forecasting the loads beyond 1994. Since the weather condition data is not available for longer than two weeks ahead, a sensitivity program is developed to produce the future temperature from the present one. Very reasonable results have been obtained for the seen (inner sample) and unseen (out-of-sample or loads of target years) data. In this study, total system load forecast reflecting current and future trends, tempered with good judgment which is the key to all planning, indeed financial success is carried out for 9 power utilities in Japan. The obtained results of this study will be useful for other country's utilities.

    AB - In this paper, a recurrent neural network (RNN) is applied for long-term load forecasting. The proposed RNN is trained with the past 20 years (1975-1994) of actual data and is tested for target years (1995-1997, 2000, and 2005). In addition to the target year load forecasting, a sliding window training method is proposed for continuous retraining of the RNN. The actual data of 9 Japanese power utilities is used for forecasting the loads of 1975 to 1994. However, forecasted data is applied for forecasting the loads beyond 1994. Since the weather condition data is not available for longer than two weeks ahead, a sensitivity program is developed to produce the future temperature from the present one. Very reasonable results have been obtained for the seen (inner sample) and unseen (out-of-sample or loads of target years) data. In this study, total system load forecast reflecting current and future trends, tempered with good judgment which is the key to all planning, indeed financial success is carried out for 9 power utilities in Japan. The obtained results of this study will be useful for other country's utilities.

    UR - http://www.scopus.com/inward/record.url?scp=0031357766&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0031357766&partnerID=8YFLogxK

    M3 - Conference contribution

    AN - SCOPUS:0031357766

    VL - 2

    SP - 895

    EP - 898

    BT - Proceedings of the Universities Power Engineering Conference

    PB - Technological Educational Institute

    CY - Iraklio, Greece

    ER -