Elektr energiyasi iste’molini uzoq muddatli prognozlash uchun dastlabki ma’lumotlarni tasdiqlash

Abstract

Relevance: modern developments in the electric power industry are accompanied by increasing uncertainty in consumer behavior, a growing share of distributed energy sources, and the implementation of new digital and intelligent technologies. Under these conditions, long-term forecasting of electricity demand becomes a critical tool for strategic planning, substantiation of investment decisions, design of new power systems, and formation of a sustainable energy development strategy. The accuracy of the forecast directly depends on the quality of the initial data used in analytical and modeling procedures. Actual consumption data obtained from automated metering systems often contain outliers, gaps, and systematic distortions caused by both technical and organizational factors. Neglecting the data verification stage can lead to distortions in predictive models, which, in turn, results in errors in assessing capacity needs, energy production and supply volumes, as well as in grid infrastructure calculations. Therefore, verification of initial data is considered the most important stage of long-term demand forecasting, ensuring the reliability and informational completeness of calculations. The application of robust statistical methods, such as interquartile range analysis, allows for a significant improvement in the credibility of forecast scenarios and minimizes the risks of erroneous strategic decisions in the electric power sector.

Aim: to verify the initial data for use in the development of mathematical models for long-term electricity consumption forecasting.

Methods: there was used the interquartile range statistical method.

Results: the application of the interquartile range (IQR) method revealed no outliers in the initial data on electricity consumption, temperature, population, and GDP. This indicates high quality and stability of the data, providing a reliable foundation for developing long-term energy consumption forecasting models.