Very Short Term Load Forecasting Using Hybrid Regression and Interval Type -1 Fuzzy Inference

3rd Annual Applied Science and Engineering Conference

工业技术;自然科学

Jamaaluddin, J.^1,2 ; Robandi, I.^1

Institut Teknologi Sepuluh Nopember (ITS), Surabaya, Jawa Timur

60111, Indonesia^1

Program Studi Teknik Elektro, Universitas Muhammadiyah Sidoarjo, Jl. Raya Gelam, Candi, Sidoarjo, Jawa Timur, Indonesia^2

关键词: Distribution systems;    Electricity-consumption;    Fuzzy inference systems;    Generation systems;    Load forecasting;    Regression method;    Short term load forecasting;    Transmission systems;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/434/1/012209/pdf
DOI  :  10.1088/1757-899X/434/1/012209

来源: IOP

【 摘 要 】

The growth of electricity consumption in this world is getting higher. The operation of the electric power starts from the generation system, Transmission system, and distribution system up to the load. All systems must be well integrated. Power generation settings should be appropriate. Therefore, load forecasting is important to do in generation system so that it is not too high from the existing load. There are two kinds of load forecasting; Short Term and Very Short Term. The very short term load forecasting is to forecast the load amount in every 30 minutes on one day before the day of loading. This research aims to discuss very short term load forecasting which uses hybrid regression method in the primary data of its loading history forecasting with Interval Type - 1 Fuzzy Inference System (IT-1 FIS). The finding indicates that the forecasting in 2015 obtained error of 0,9558%, and 1,4226% in 2016.