Electric Energy Consumption Modes Forecasting and Management for Gas Industry Enterprises Based on Artificial Intelligence Methods

This paper shows the electrical load forecast problem solution for gas industry enterprises, taking into account the comparative analysis of forecasting methods. The obtained information is an effective tool for influencing the optimization of the tariff policy and further corrective well-timed actions of personnel for improving contractual work and minimizing costs for consumed electricity. Artificial neural network models are developed taking into account electric energy consumption nature and parameters. This paper presents the dependences of changes in the coefficients of seasonality, cyclicity, and irregularity for the electric load at electricity metering points. The problem of forecasting electrical loads and predicting their accuracy is shown and solved in the context of the gas industry enterprises. The characteristics and architectures of multiple artificial neural networks comparative analysis is given.