<![CDATA[In electric power operations, the use of energy often leads to decreased energy efficiency. Some of the main causes of power load instability are due to sudden changes or shifts, which cause the fluctuation of electrical energy to become imbalanced. If this instability persists over a long period, it poses a risk to the electrical system and can damage components. To prevent this, a load forecasting method is employed to determine the loading for a specific period. Electric energy forecasting is frequently used to determine the loading parameters that will occur in the future. This is done to maintain the quality of the electric power and ensure adequate energy supply. To provide accurate electric energy forecasting, several methods are often used, one of which is the artificial neural network (ANN). An ANN can use parameters not only derived from electrical measurements and main loading but also allows for non-linear computations. There are several ANN methods, one of which is Long Short-Term Memory (LSTM). Long Short-Term Memory is used when the input parameters are non-linear or uncorrelated. Therefore, to obtain maximum forecasting results, electric energy forecasting using Long Short-Term Memory was employed for a high-rise building. The research results then showed that a learning rate of 0.1 yielded an MAPE (Mean Absolute Percentage Error) of 6.41% and an RMSE (Root Mean Square Error) of 9.4.]]>
Copyrights © 2025
