OPTIMIZATION OF THE STRUCTURE OF THE ENERGY CONSUMPTION FORECASTING SYSTEM WITH ATYPICAL ENERGY CONSUMPTION PATTERNS

Abstract

The creation of an intelligent energy consumption forecasting device for consumers with atypical
energy consumption is considered, depending on the required forecast accuracy, taking into account, in
addition to the target parameters of the power grid (P, Q), technological processes of enterprises, influencing
factors: socio-economic (hour of the day; day of the week; ordinal number of the day in the year;
sign of a holiday or mass events d); meteorological: (wind-cold index). The model refers to intelligent devices for adaptive forecasting of power consumption modes of the electric grid based on a multilayer
neural network. The article is devoted to the choice of the optimal architecture of the neural network (NN)
and the method of its training, providing forecasting with the least error. A multifactional model of power
consumption based on a multilayer NN has been synthesized and tested. Within the framework of the conducted
research, an NN model was built describing the architecture of a cyber-physical system (CFS) for
forecasting power consumption. It has been established that for each consumer, due to significant differences
in the nature of energy consumption, it is necessary to experimentally select network parameters in
order to achieve a minimum prediction error. It is shown that with atypical power consumption, i.e., not
repeated over time periods (hour, day, week, etc.), artificial intelligence and deep machine learning methods
are an effective tool for solving poorly formalized or non-formalized tasks. The developed model has
acceptable accuracy (MSE deviation up to 15%). To increase the accuracy of the forecast, it is necessary
to carry out a regular refinement of the model and adjust it to the actual situation, taking into account new
additive factors affecting the electricity consumption curve. The possibility of using this device in the technological
management systems of regional grid companies, which forms the basis of a hierarchical automated
information measuring system for monitoring and accounting for electricity, by accounting and
forecasting the active and reactive power of electric consumers