The dissertation work developed models, methods and information technology for analyzing energy conservation at metallurgical enterprises, which will provide a solution to the scientific and practical problem of managing energy conservation processes, taking into account the dynamics of energy consumption, economic and environmental risks.
Main results of the work.
1. For the first time, a concept for increasing energy conservation at a metallurgical enterprise was proposed based on information technology for managing energy conservation processes, on the basis of which a sequence of tasks was developed, which includes the technology of express energy audit, reducing the error in predicting energy consumption, visualization of deviations from regulatory indicators, which makes it possible to prevent economic and environmental risks.
2. As a result of the express energy audit based on data on hourly electricity consumption of "DNIPROSPETSSTAL" and day-night price zones, it was shown how to obtain an economic effect from saving energy consumption, taking into account price zones, and appropriate recommendations were given.
3. When building forecasting models, the work compared several neural network architectures and selected the two best ones that provide the highest accuracy in forecasting energy consumption for the day ahead - a nonlinear autoregression model with the Levenberg-Marquardt learning algorithm and an LSTM network. The economic effect of this work is formed by reducing the overall error in forecasting electricity consumption.
4. According to the results of the study of the influence of meteorological factors (temperature, illumination) on the energy consumption processes of metallurgical enterprises, it was concluded that in the winter period of the year, energy consumption significantly depends on temperature and does not depend at all on cloudiness. Regarding the summer period, it can also be stated that energy consumption depends on temperature, but it is much weaker. In connection with climate change (annual temperature increase), it is recommended to monitor temperature changes as a means of saving when heating furnaces.
5. Analysis of energy consumption and based on recurrent analysis of the metallurgical enterprise by the winter-summer seasons showed that the summer period is characterized by greater predictability, as well as a significantly higher delay indicator, which characterizes the average time that the system can spend in a more or less unchanged state.
6. Improving the information technology for monitoring environmental and economic risks of metallurgical enterprises by developing an infological model of the analytical panel will allow for a prompt response to deviations from regulatory indicators, thereby increasing environmental efficiency, ensuring cost-effective operation and promoting the development of harmonious interaction between the community and enterprises of the metallurgical complex. Monitoring of environmental and economic risks will allow for more accurate initial information for making management decisions regarding their implementation under conditions of different equipment loads due to different production intensities.
