Low Emission Building Control with Zero-Shot Reinforcement Learning

Heating and cooling systems in buildings account for 31\% of global energy use, much of which are regulated by Rule Based Controllers (RBCs) that neither maximise energy efficiency nor minimise emissions by interacting optimally with the grid. Control via Reinforcement Learning (RL) has been shown to significantly improve building energy efficiency, but existing solutions require access to building-specific simulators or data that cannot be expected for every building in the world. In response, we show it is possible to obtain emission-reducing policies without such knowledge a priori--a paradigm we call zero-shot building control. We combine ideas from system identification and model-based RL to create PEARL (Probabilistic Emission-Abating Reinforcement Learning) and show that a short period of active exploration is all that is required to build a performant model. In experiments across three varied building energy simulations, we show PEARL outperforms an existing RBC once, and popular RL baselines in all cases, reducing building emissions by as much as 31\% whilst maintaining thermal comfort. Our source code is available online via https://enjeeneer.io/projects/pearl .

Zero-Shot Building Control

Heating and cooling systems in buildings account for 31% of global energy use, much of which are regulated by Rule Based Controllers (RBCs) that neither maximise energy efficiency nor minimise emissions by interacting optimally with the grid. Control via Reinforcement Learning (RL) has been shown to significantly improve building energy efficiency, but existing solutions require pre-training in simulators that are prohibitively expensive to obtain for every building in the world. In response, we show it is possible to perform safe, zero-shot control of buildings by combining ideas from system identification and model-based RL. We call this combination PEARL (Probabilistic Emission-Abating Reinforcement Learning) and show it reduces emissions without pre-training, needing only a three hour commissioning period. In experiments across three varied building energy simulations, we show PEARL outperforms an existing RBC once, and popular RL baselines in all cases, reducing building emissions by as much as 31% whilst maintaining thermal comfort.