B2RL: An open-source Dataset for Building Batch Reinforcement Learning

被引:3
作者
Liu, Hsin-Yu [1 ]
Fu, Xiaohan [1 ]
Balaji, Bharathan [2 ]
Gupta, Rajesh [1 ]
Hong, Dezhi [2 ]
机构
[1] Univ Calif San Diego, La Jolla, CA 92093 USA
[2] Amazon, New York, NY USA
来源
PROCEEDINGS OF THE 2022 THE 9TH ACM INTERNATIONAL CONFERENCE ON SYSTEMS FOR ENERGY-EFFICIENT BUILDINGS, CITIES, AND TRANSPORTATION, BUILDSYS 2022 | 2022年
关键词
COMFORT;
D O I
10.1145/3563357.3566164
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Batch reinforcement learning (BRL) is an emerging research area in the RL community. It learns exclusively from static datasets (i.e. replay buffers) without interaction with the environment. In the offline settings, existing replay experiences are used as prior knowledge for BRL models to find the optimal policy. Thus, generating replay buffers is crucial for BRL model benchmark. In our B2RL (Building Batch RL) dataset, we collected real-world data from our building management systems, as well as buffers generated by several behavioral policies in simulation environments. We believe it could help building experts on BRL research. To the best of our knowledge, we are the first to open-source building datasets for the purpose of BRL learning.
引用
收藏
页码:462 / 465
页数:4
相关论文
共 30 条
[1]  
Balaji Bharathan, 2013, BuildSys, P1
[2]  
Brockman G, 2016, Arxiv, DOI arXiv:1606.01540
[3]   EnergyPlus: creating a new-generation building energy simulation program [J].
Crawley, DB ;
Lawrie, LK ;
Winkelmann, FC ;
Buhl, WF ;
Huang, YJ ;
Pedersen, CO ;
Strand, RK ;
Liesen, RJ ;
Fisher, DE ;
Witte, MJ ;
Glazer, J .
ENERGY AND BUILDINGS, 2001, 33 (04) :319-331
[4]  
de Dear R.J, 1998, ASHRAE Trans., V104, P1141
[5]   MB2C: Model-Based Deep Reinforcement Learning for Multi-zone Building Control [J].
Ding, Xianzhong ;
Du, Wan ;
Cerpa, Alberto E. .
PROCEEDINGS OF THE 7TH ACM INTERNATIONAL CONFERENCE ON SYSTEMS FOR ENERGY-EFFICIENT BUILDINGS, CITIES, AND TRANSPORTATION, BUILDSYS 2020, 2020, :50-59
[6]  
Fanger P. O., 1970, Thermal comfort. Analysis and applications in environmental engineering.
[7]   Beobench: A Toolkit for Unified Access to Building Simulations for Reinforcement Learning [J].
Findeis, Arduin ;
Kazhamiaka, Fiodar ;
Jeen, Scott ;
Keshav, Srinivasan .
PROCEEDINGS OF THE 2022 THE THIRTEENTH ACM INTERNATIONAL CONFERENCE ON FUTURE ENERGY SYSTEMS, E-ENERGY 2022, 2022, :374-382
[8]  
Fu J., 2020, arXiv, DOI 10.48550/arXiv.2004.07219
[9]  
Fujimoto S, 2021, ADV NEUR IN, V34
[10]  
Fujimoto S, 2019, PR MACH LEARN RES, V97