<![CDATA[The availability of adequate and targeted electricity can accelerate the development of the region. Predicting the demand for electricity energy with a high degree of accuracy can help plan the production and distribution of electricity efficiently and efficiently. The study predicted the size of medium-term electricity consumption in the PLN (Persero) Area of Padang power system to help decision-making in generation and distribution planning. Backpropagation algorithms are an effective method that can be applied with a high degree of accuracy. Factors that can influence electricity consumption are used as input: economic growth, consumer growth, population and installed power. Meanwhile, electricity consumption rates are used as outputs. The data used are historical data of electricity consumption, customer growth, GDP over the period 2012-2017. Data 2012-2016 used as training data, data 2017 used as test data. Simulation and data processing using MATLAB software support. After processing and data analysis, the best architecture 4-2-1 with learning rate = 0.1, momentum = 0.9 yields the best prediction accuracy of 94.81 % and MAPE value = 5.1878. Proper prediction can help plan the supply of electricity needs. Backpropagation algorithms with correct parameters produce predictions with the best accuracy.]]>
Copyrights © 2022
