AI and building energy
Learning objectives
By the end of the course, participants will be able to:
- Identify a building’s energy data sources and their limits
- Prepare consumption time series
- Build a consumption forecasting model and measure its error
- Detect operational anomalies
- Discuss the use of these models for renovation and building management, with their limits
Audience
Engineering school or master’s students (civil engineering, energy, building services).
Prerequisites
Basic building physics; basics of Python or another scientific computing language.
Day-by-day programme
| Session | Content |
|---|---|
| Day 1 morning | Building energy data: meters, sub-metering, sensors, weather data; quality and limits |
| Day 1 afternoon | Preparing time series: cleaning, resampling, explanatory variables (weather, occupancy, calendar) |
| Day 2 morning | Consumption forecasting: regression and machine learning models, validation, comparison with a simplified physical approach |
| Day 2 afternoon | Anomaly detection and decision support; mini-project on a building dataset, presentation |

