CityLearn for Human-Centered Grid Flexibility
summary
CityLearn v3 is an open-source simulator now in the LF Energy ecosystem that models HVAC, PV, batteries, EV charging and occupant-centric control across heterogeneous buildings and districts. A controller trained on a 17-building zero-net-energy community in California learns occupant preferences and shifts heating/cooling to provide 60% improvement in district flexibility, often reducing the need for extra batteries. Current deployments include EV fleet trials in Portugal and cold-climate experiments in Vermont; the platform exposes grid signals and building observations for coordinated controls.
AI-written summary · about
chapters (6)
- 0:00Introduction to CityLearn
- 2:05Human-centered grid flexibility
- 3:29Zero net energy community case
- 5:18Version 3 updates and heterogeneity
- 7:03Global coordination efforts
- 8:33Q&A on energy arbitrage
related talks
01h 28mTools and data to decarbonize cities: Citiwatts, CoolLIFE & an EU-27 Mapping of Financing Schemes
23mSustainable Cities
21mPublic services Goes Digital
40mSmart Grid: How IoT fights climate change
31mLessons Learnt Visualising Real-World Cities in 3D
11mUXers
42mHow Open Source...
43mMore Renewable Energy Into the Power Grid with Open Source