We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.SY

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Systems and Control

Title: Deep Reinforcement Learning for Smart Building Energy Management: A Survey

Abstract: Global buildings consumed 30% of total energy and generated 28% of total carbon emission in 2018, which leads to economic and environmental concerns. Therefore, it is of great significance to reduce energy consumption, energy cost and carbon emission of buildings while maintaining user comfort. To this end, several challenges have to be addressed. Firstly, it is very challenging to develop a building thermal dynamics model that is both accurate and efficient enough for building control. Secondly, there are many kinds of uncertainties. Thirdly, there are many spatially and temporally operational constraints. Fourthly, building energy optimization problems may have extremely large solution spaces, which can not be solved in real-time by traditional methods. Fifthly, traditional building energy management methods have respective applicable premises, which means that they have low versatility when confronted with varying building environments. As a general artificial intelligence technology, deep reinforcement learning (DRL) has the potential of addressing the above challenges. Thus, this paper presents a comprehensive literature review on DRL for smart building energy management (SBEM). To be specific, we first introduce the fundamentals of DRL and provide the classification of DRL methods used in existing works related to SBEM. Then, we review the applications of DRL in a single building energy subsystem, multiple energy subsystems of buildings, and building microgrids, respectively. Furthermore, we identify the unsolved issues and point out the possible research directions of applying DRL. Finally, we summarize the lessons learned from this survey.
Comments: 17 pages, 8 figures
Subjects: Systems and Control (eess.SY); Signal Processing (eess.SP)
Cite as: arXiv:2008.05074 [eess.SY]
  (or arXiv:2008.05074v2 [eess.SY] for this version)

Submission history

From: Liang Yu [view email]
[v1] Wed, 12 Aug 2020 02:46:48 GMT (2653kb)
[v2] Mon, 9 Nov 2020 14:29:01 GMT (2517kb)
[v3] Wed, 22 Sep 2021 14:01:23 GMT (3407kb)

Link back to: arXiv, form interface, contact.