A Simplified Building Control Environment with a Reinforcement Learning Application

Repository to accompany "A Simplied Building Controls Environment with a Reinforcement Learning Application" paper submitted to IBPSA Rome 2019 Conference. Paper authors: Vasken Dermardiros, Scott Bucking and Andreas K. Athienitis.

Please consult the following links to be more accustomed to how OpenAI Gym and Tensorforce work:

• OpenAI Documentation: https://gym.openai.com/docs
• Creating a new OpenAI Env: https://github.com/openai/gym/tree/master/gym/envs#how-to-create-new-environments-for-gym
• OpenAI Env Wiki: https://github.com/openai/gym/wiki/Environments
• Simple example: https://github.com/MartinThoma/banana-gym
• OpenAI Baselines: https://github.com/openai/baselines
• Tensorforce: https://github.com/tensorforce/tensorforce

To install, go to gym-BuildingControls folder and run the command:

pip install -e .

Cite SimplifiedBldgControlEnv

Please cite the framework as follows:

@inproceedings{dermardiros_ibpsa_2019,
  author       = {Dermardiros, Vasken and Bucking, Scott and Athienitis, Andreas K.},
  title        = {A {Simplified} {Building} {Control} {Environment} with a {Reinforcement} {Learning} {Application}},
  booktitle    = {Proceedings of {IBPSA} 2019, {Rome}, {Italy}},
  url          = {https://github.com/vderm/SimplifiedBldgControlEnv},
  year         = {2019}
}