The research aims to predict the demand for electric energy through the use of artificial intelligence techniques to predict the future demand for energy in the diesel station north of Al-Amarah to generate electric power. Which suffers from the problem of uncertainty in the demand for future energy, due to weakness in knowing future loads, which leads to lack of planning Correct scheduling of productive operations and management of maintenance work, and thus a decrease in the reliability of the system.

Especially since it is boosting the capacity of the national network in the main roles (continuous, peak, and reserve), and this made it more difficult to predict the demand for future loads with high accuracy. As artificial intelligence techniques and traditional techniques were used in the completion of the study. As the researcher used the traditional model ARIMA in the program SPSS and the neural group data processing model (GMDH) of the type of neural network (ANN), and the neural network model of the feed-forward backprop type to predict energy. Demand based on historical data of daily electrical loads for the period from 1/1/2019 to 31/12/ 2021, as the data volume reached (1096) per day.

The research reached a set of results, the most important of which is obtaining prediction values for the future demand for energy for a month starting from 1/1/2022 to 31/1/2022 for the station, as the neural network model GMDH gave the lowest value of the mean absolute relative error (MAPE) of 0.0567. While the feed-forward backprop method gave a mean absolute relative error (MAPE) of 0.0648, and the traditional ARIMA model gave a mean absolute relative error (MAPE) of 0.0654, as the obtained results show the effectiveness of the GMDH-type neural network model in prediction.