Abstract:
Systematic research into the operations of the regional power system aimed at improving the efficiency of energy complex management, taking into account the contribution of utilized resources, is fundamentally impossible without the enhancement of mathematical models and methods for their identification based on statistical data.
This article presents the results of an analysis of a well-known mathematical description of the functioning of the regional power system, highlighting significant shortcomings that negatively impact both the reliability of assessments of key performance indicators of the energy complex and the accuracy of forecasts made based on the constructed model.
The study examines and systematizes various three-factor regression models and covariance-stationary time series models based on linear and nonlinear regression into three main groups. Algorithms for numerical methods of least squares estimation of the parameters of these models based on observational results are described.
Results of mathematical modeling of the dynamics of energy system output based on statistical data published in the annual reports of regional ministries and energy companies are provided. A statistical analysis of the obtained results is conducted. A comparative analysis of the developed mathematical models based on forecast error assessment allowed for the selection of the most effective mathematical model with minimal forecasting error from the considered set of models over a time period ranging from one to five years.
Keywords:regional energy system, multiplicative power-law production function, factor elasticity, covariance-stationary time series models, autoregressive models, statistical analysis
Fulltext:
PDF file (1079 kB)
The content is published under the terms of the Creative Commons Attribution 4.0 International License
