PERAMALAN KEBUTUHAN ENERGI LISTRIK JANGKA PANJANG SISTEM KETENAGALISTRIKAN DKI JAKARTA MENGGUNAKAN METODE FUZZY LOGIC DENGAN JARINGAN SARAF TIRUAN
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Marpaung, Frintis Septa
Arifin, Zainal
Arifin, Zainal
Subject
Teknik Elektro
Datestamp
2022-12-02 07:30:41
Abstract :
Availability of electrical energy in appropriate quantities is an absolute must to maintain human life. Moreover, with the development and penetration of micro- hydro power plants, photo voltage and other renewable energies, the need for long- term load forecasting in the distribution system is increasingly needed. especially in energy planning because proper planning will have an effect on more targeted implementation. Method for long term load forecasting is fundamentally centered on fuzzy logic compared with artificial neural network. The long term relations in a time series data of electricity load demand, economic and demography are taken into account using Matlab. Precisely, publicly available data of fourteen years from 2006 to 2019 have been collected to train and validate the model. .From the data, the results of forecasting with fuzzy logic are not much different from the results of forecasting by artificial neural networks with an average error of 2% for Sugeno fuzzy logic and 2% for Mamdani fuzzy logic and 3% for forecasting with artificial neural network algorithm methods. Bayesian Regularization and 2% for Levenberg Marquard algorithm.By comparing the results of forecasting for 5 years from the four methods, it is concluded Forecasting long-term electricity demand using artificial neural network and fuzzy logic methods is more accurate than RUPTL because it produces a smaller error margin of 2-3% compared to the RUPTL error margin which is more than 10%.
Availability of electrical energy in appropriate quantities is an absolute must to maintain human life. Moreover, with the development and penetration of micro- hydro power plants, photo voltage and other renewable energies, the need for long- term load forecasting in the distribution system is increasingly needed. especially in energy planning because proper planning will have an effect on more targeted implementation. Method for long term load forecasting is fundamentally centered on fuzzy logic compared with artificial neural network. The long term relations in a time series data of electricity load demand, economic and demography are taken into account using Matlab. Precisely, publicly available data of fourteen years from 2006 to 2019 have been collected to train and validate the model. .From the data, the results of forecasting with fuzzy logic are not much different from the results of forecasting by artificial neural networks with an average error of 2% for Sugeno fuzzy logic and 2% for Mamdani fuzzy logic and 3% for forecasting with artificial neural network algorithm methods. Bayesian Regularization and 2% for Levenberg Marquard algorithm.By comparing the results of forecasting for 5 years from the four methods, it is concluded Forecasting long-term electricity demand using artificial neural network and fuzzy logic methods is more accurate than RUPTL because it produces a smaller error margin of 2-3% compared to the RUPTL error margin which is more than 10%.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
