Penerapan Metode Monte Carlo Untuk Peramalan Beban Puncak
Listrik Jangka Pendek
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Pratomo, Thareq Tito
Siregar, Riki Ruli Affandi
Kuswardani, Dwina
Siregar, Riki Ruli Affandi
Kuswardani, Dwina
Subject
Teknik Informatika
Datestamp
2022-09-23 03:15:11
Abstract :
The use of electrical energy loads will arise problems if the power generated is greater than the demand on the side of the load, then there will be energy waste. Conversely, if the power generated is smaller than the demand on the side of the load, there will be a blackout. Short-term peak load forecasting at PT. PLN (Persero) East Java has been using the manual input system through spreadsheet. The Monte Carlo method can be used to forecast short-term peak loads using historical data at PT. PLN (Persero) East Java. There are 3 (three) series of experimental simulations this time which aim to compare the results of forecasting on each sample. The forecasting result is the value of the load for the next 7 (seven) days so that there are 7 (seven) forecasting times in each sample of the simulation data. A large number of data samples in the Monte Carlo calculation process affects the accuracy of forecasting. The level of accuracy in forecasting using the first simulation, namely 7 load data with 20 random number samples reaching 91.29% with MAPE value of 8.71%. By using the second simulation of 14 load data with 40 samples of random numbers obtained the results of MAPE values of 5.73% with an accuracy of 94.27%. In the third simulation, there were 21 load data with 60 random number samples resulting in MAPE values of 4.17% with an accuracy of 95.83%.
The use of electrical energy loads will arise problems if the power generated is greater than the demand on the side of the load, then there will be energy waste. Conversely, if the power generated is smaller than the demand on the side of the load, there will be a blackout. Short-term peak load forecasting at PT. PLN (Persero) East Java has been using the manual input system through spreadsheet. The Monte Carlo method can be used to forecast short-term peak loads using historical data at PT. PLN (Persero) East Java. There are 3 (three) series of experimental simulations this time which aim to compare the results of forecasting on each sample. The forecasting result is the value of the load for the next 7 (seven) days so that there are 7 (seven) forecasting times in each sample of the simulation data. A large number of data samples in the Monte Carlo calculation process affects the accuracy of forecasting. The level of accuracy in forecasting using the first simulation, namely 7 load data with 20 random number samples reaching 91.29% with MAPE value of 8.71%. By using the second simulation of 14 load data with 40 samples of random numbers obtained the results of MAPE values of 5.73% with an accuracy of 94.27%. In the third simulation, there were 21 load data with 60 random number samples resulting in MAPE values of 4.17% with an accuracy of 95.83%.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
