INTELLIGENT FORECASTING OF RENEWABLE GENERATION FOR DEFICIT REDUCTION AND GRID STABILITY

Abstract

The increasing penetration of variable renewable energy sources creates new challenges for regional power systems characterized by structural electricity deficits and dependence on external power transfers. This paper develops an intelligent forecasting framework for wind and solar power generation aimed at supporting power balance assessment and grid stability analysis in the Almaty region of Kazakhstan. The proposed approach combines historical meteorological data, renewable generation profiles, and machine-learning algorithms to forecast variable renewable energy output under different operating conditions. Forecasting models based on time-series analysis and machine-learning techniques are evaluated using standard accuracy indicators, including RMSE, MAE, and MAPE. The forecasted renewable generation is further integrated into a regional power balance assessment to estimate its potential contribution to reducing electricity deficit and supporting operational planning. The results demonstrate that intelligent forecasting can improve the predictability of renewable generation, reduce uncertainty in dispatch decisions, and provide a data-driven basis for renewable energy integration in deficit prone power systems. The study contributes to the development of practical forecasting tools for emerging power systems with growing shares of renewable energy. Keywords: intelligent forecasting, Renewable energy forecasting, variable renewable energy, wind power forecasting, machine learning, artificial intelligence.