INHALT
Electricity demand forecasting is important for planning and expanding facilities in the power sector. Accurate forecasts can save operation and maintenance costs, increase the reliability of the power supply system and make sound decisions for future development. As the population grows and the ever-increasing demand for energy not only from households but also from businesses, traditional forecasting methods often find it difficult to respond to an increasingly complex and constantly demanding and evolving environment.
This thesis will examine different methods and data analysis techniques that will help to better forecast electricity demand. Extracting this information will have multiple benefits, especially for those who produce energy. Furthermore, a series of statistical tools, such as ARIMA and regression, will be explored.
In order to be a substantial improvement in the forecasts, the study will also consider various external factors, such as temperature fluctuations, official holidays, economic circumstances, geopolitical disturbances, etc. The data should be properly cleaned and normalized to produce the correct forecast before entering the model. The studied improvement models will be compared and evaluated based on the available criteria.
This research offers valuable insights for energy providers, policymakers, and grid operators by comparing different forecasting approaches. More accurate forecasts can help optimize energy production, reduce waste and support the integration of renewable energy sources into the grid.
The findings aim to contribute to smarter and more efficient energy management, making forecasting electricity demand more reliable and stable in an increasingly dynamic energy landscape. |
|