The issue of using a long-term prehistory to improve the accuracy of short-term forecasting the total electric load of the power system is considered. Information on hourly loads of regional energy systems and the main factors that affect them is stored in the developed database. The special control program can display information in graphically. Using the program for processing

Based on the research, the article presents three algorithms that allow to select from the overall electrical load (OEL) of the power system technological and temperature components in each hour of the daily schedule, which provides greater accuracy of short-term forecasting (STF) OEL of the power system. Calculations by three algorithms were performed according to Kyivenerho. The analysis of reading temperature sensors on four sources from the point of view of possibility of their application at STF is carried out. References 6, figures 3, tables 3.

No abstract

The paper proposes the architecture of deep learning neural network for short-term nodal electrical load forecasting. The neural network combines the recurrent module LSTM (Long short-term memory) and the multilayer perceptron on the top. Input and output of the network connected with shortcut connection. In multilayer perceptron used scaled exponential linear unit (SELU) function as a nonlinear transformation in hidden neurons. A comparative analysis of two approaches to the short-term prediction of node loadings of the grid is conducted. In the first approach, a separate model based on the artificial neural network eResNet is built for each load node. In the second approach, vector prediction of the values of the nodal load is performed using the proposed neural network. The second approach makes it possible to exploit the relationship between the loads in the nodes and reduce the number of computational operations required to build the model, especially at a large number of nodes. Recurrent network showed slightly better result when forecasting horizon was 24 hours, but eResNet showed more accurate forecast with longer horizons. References 16, figure 1, tables 3.

No abstract