Short-term Load Forecasting Using Statistical Methods for the Central Power System of Mongolia

A feature of the power system is that all the processes of production, transmission and distribution occur simultaneously. This complicated and continuous process relies on power system management. For effective management and improve reliability, it is necessary to study the characteristic of electricity consumption. Forecasting of the power demand allows to optimize the power distribution. Therefore, forecasting is given a lot of attention in the energy industry. Purpose is to forecast the daily load curve for weekdays, considering the influence of meteorological factors. This study is carried out by the method of statistical analysis. The historical data of 2021 on energy consumption, temperature and outdoor air humidity of the Central energy system of Mongolia were used as initial inputs. According to the method of statistical analysis, daily load curves were constructed with an absolute percentage error of 2.68%. To confirm this result, an analysis was carried out using the method of autoregressive integrated moving average. After considering the meteorological factor, the absolute percentage error decreased down to 2.26%. The proposed method proved capable to forecast load curves with appropriate accuracy.