PRAKIRAAN KAPASITAS PENGHANTAR TRANSMISI TERHADAP BEBAN LISTRIK JANGKA PANJANG DENGAN METODE JARINGAN SYARAF TIRUAN (JST)
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Delpiero, Joshua Fransisco
Siregar, Masbah R.T
Qosim, Muchaman Nur
Siregar, Masbah R.T
Qosim, Muchaman Nur
Subject
Teknik Elektro
Datestamp
2022-09-28 03:18:40
Abstract :
Electrical load forecasting is a method for predicting the increase or trend of load from an area. Long-term electrical load forecasts are used to plan and as a reference for the electric power system on energy availability and equipment capacity. The capacity of the line equipment in a transmission network depends on the value of the load being carried and has a limit on the ability to overload. The load forecast is used to ensure the value of maintenance the Madura Island cable conductor is not overcapacity. As a prediction method, there are many variations of long-term electrical load prediction methods. Artificial Neural Network (ANN) was chosen as a long-term electrical load prediction method. By using the Neural Network Method or neural network is expected to have a high accuracy value by looking at the error value (MSE) of each prediction method. The method used is traingd and traincgb training as a pattern of variations and differences in hidden layer values. By using the traincgb training method with the smallest average error value (MSE) is 0.0000000682 so that the traincgb training with the number of hidden layers is 5 used as a forecast method. So that the predicted load is obtained in 2021 to 2027. With the limit value of Madura conductor cable overcapacity in September 2026 with a load of 293,2979 MW where the current installed Madura conductor limit value is 290 MW. With these forecasts, it is hoped that the prediction of the electrical load can be used as a comparison and reference
Electrical load forecasting is a method for predicting the increase or trend of load from an area. Long-term electrical load forecasts are used to plan and as a reference for the electric power system on energy availability and equipment capacity. The capacity of the line equipment in a transmission network depends on the value of the load being carried and has a limit on the ability to overload. The load forecast is used to ensure the value of maintenance the Madura Island cable conductor is not overcapacity. As a prediction method, there are many variations of long-term electrical load prediction methods. Artificial Neural Network (ANN) was chosen as a long-term electrical load prediction method. By using the Neural Network Method or neural network is expected to have a high accuracy value by looking at the error value (MSE) of each prediction method. The method used is traingd and traincgb training as a pattern of variations and differences in hidden layer values. By using the traincgb training method with the smallest average error value (MSE) is 0.0000000682 so that the traincgb training with the number of hidden layers is 5 used as a forecast method. So that the predicted load is obtained in 2021 to 2027. With the limit value of Madura conductor cable overcapacity in September 2026 with a load of 293,2979 MW where the current installed Madura conductor limit value is 290 MW. With these forecasts, it is hoped that the prediction of the electrical load can be used as a comparison and reference
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
