Open AccessTime Difference Calculation Settings for Very Short Term Electric Load Forecasting Using Interval Type-1 Fuzzy Inference System (IT-1 FIS)
Author(s) -
Jamaaluddin Jamaaluddin,
Imam Robandi,
Izza Anshory
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/874/1/012011
Subject(s) - thursday , term (time) , electricity , interval (graph theory) , statistics , value (mathematics) , operations research , econometrics , computer science , operations management , mathematics , economics , engineering , philosophy , linguistics , physics , quantum mechanics , combinatorics , electrical engineering
Very short term electricity load forecasting (VSTLF) is done the day before forecasting every 30 minutes. This forecasting is imperative about the planning of operating plants, which are carried out in maintenance and related to economic considerations. This forecasting refers to the loading of electricity in a particular area and time. Very short-term electricity loading in the past few years has been used to predict the electricity load shortly. One method used for forecasting very short loads is using IT-1 FIS. In IT-1-FIS, the dataset is made by looking at the imposition of electricity on certain days at certain hours of the previous year. The data that calculate time difference forecast Friday 4th used Friday 3rd, 2nd, and 1st at the same hour and month of the past year. In this study, it proposed that the calculation of Time difference forecasting the 4th Friday be used data Thursday 4th, 4th Wednesday, 4th Tuesday. After entering into IT-1-FIS, the forecasting error result is 0,6123 %. This value is higher than if you use the time difference on Friday 3rd, 2nd, and 1st, which has a forecast error value of 0,4695 %. So it can be concluded that the time difference setting is very influential in the process of calculating forecasts, and the use of time difference data with the use of Friday data 3rd, 2nd, and 1st is better.