Real-world challenges for multi-agent reinforcement learning ingrid-interactive buildings

Building upon prior research that highlighted the need for standardizing environments for building controlresearch, and inspired by recently introduced challenges for real life reinforcement learning (RL) control, herewe propose a non-exhaustive set of nine real world challenges for RL control in grid-interactive buildings(GIBs). We argue that research in this area should be expressed in this framework in addition to providing astandardized environment for repeatability. Advanced controllers such as model predictive control (MPC) andRL control have both advantages and disadvantages that prevent them from being implemented in real worldproblems. Comparisons between the two are rare, and often biased. By focusing on the challenges, we caninvestigate the performance of the controllers under a variety of situations and generate a fair comparison.