In the first 18 months of the DR-RISE project, SUPSI developed two key software tools to enhance energy management: the Consumer Load Forecast Service and EV-Insights. The Consumer Load Forecast Service offers advanced residential energy demand prediction using innovative models with dynamic weighting, probabilistic hierarchical forecasting, and a fast adaptive exponential smoothed Fourier model. Built on a modular architecture with MLFlow, Celery, and time-series databases, it provides a user-friendly REST API for seamless integration. Simultaneously, EV-Insights addresses electric vehicle charging data management challenges by offering four essential services: Data Administration for organizing charging session data, Data Ingestion for integrating public datasets from multiple pilot projects, Analysis tools for visualizing charging patterns, and Forecasting capabilities using tree boosting methods to predict charging session duration, energy demand, and station-level charging requirements. Both tools aim to forecast energy usage, support demand response strategies, and provide flexible, adaptive solutions for residential and electric vehicle charging contexts.
Enhancing Energy Management: SUPSI’s Innovative Tools for Residential Demand Forecasting and EV Charging Insights
