Summary of Experimental Evaluation Of Offline Reinforcement Learning For Hvac Control in Buildings, by Jun Wang et al.
Experimental evaluation of offline reinforcement learning for HVAC control in buildings
by Jun Wang, Linyan Li, Qi Liu, Yu Yang
First submitted to arxiv on: 15 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The study explores the application of reinforcement learning (RL) techniques for dynamic heating, ventilation, and air conditioning (HVAC) control in buildings. While previous research has focused on online or off-policy scenarios, this paper investigates the feasibility and effectiveness of using RL-based HVAC controllers with purely offline datasets or trajectories. The authors comprehensively evaluate state-of-the-art offline RL algorithms through analytical and numerical studies, examining algorithm characteristics and dataset conditions. They demonstrate that certain offline RL algorithms can be effectively used to train well-performed RL-based HVAC controllers, achieving reduced violation ratios of indoor temperatures and power savings compared to baseline controllers. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This study uses special computer learning techniques called reinforcement learning to control heating, ventilation, and air conditioning systems in buildings. The goal is to make sure the temperature inside the building stays comfortable and the energy used is efficient. Right now, most research focuses on using these techniques when there’s a lot of data coming in, but this study looks at what happens when you only have old data. The researchers tested different ways of using reinforcement learning and found that some methods work better than others. They also discovered that you can use old data to train a computer program that can control the HVAC system well and save energy. |
Keywords
* Artificial intelligence * Reinforcement learning * Temperature
