Forecasting and Enhancing the Performance of the Electric Grid: A Dynamic Mode Decomposition approach

Abstract

The recent development in the field of smart meters has resulted in a large amount of electricity consumption profile which can be forecasted and analyzed for better planning of grid such as load scheduling, demand-side management, etc. In this paper, a Dynamic Mode Decomposition (DMD) for forecasting of grid profile for a specified time period is proposed. The key feature of the DMD technique is that it utilizes the past data for the prediction of future data without the need for the system model. Once the load profile is forecasted using DMD, the peak loads and base load time period are segregated. With the help of a system operator, various energy sources can be arranged for supplying peak loads. Electric vehicles (EVs) having the advantage of mobility and smart building acting as the virtual battery is considered for providing support to the grid as compared to other renewable energy sources which generally have limitations due to environmental factors. The DMD technique for forecasting the load profile of the smart grid is tested under various test scenarios. Finally, from the results, it has been proved that EVs and the smart building seem to be the most promising solution for supporting the grid during the peak load period.