Control for the People: How Machine Learning Enables Efficient HVAC Use Across Diverse Thermal Preferences
Control for the People: How Machine Learning Enables Efficient HVAC Use Across Diverse Thermal Preferences
复制标题
对人的控制：机器学习如何在不同的热偏好中实现 HVAC 的高效使用
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
T. Hoyt
中科院分区：
文献类型：
--
作者：
L. Baker;T. Hoyt
商业建筑物的居住者往往具有影响HVAC操作的受限手段。因此，节能行为改变的机会仅限于居住者直接控制的规定。居住者可以在晚上关灯，拔掉电子设备插头，以及他们可以在日常生活中采取的其他类似行动。然而，这些行为变化未能解决HVAC能耗问题，这是一种可以从居住者反馈中受益匪浅的能耗。收集热舒适偏好的系统将为节能创造机会，但如何激励居住者表达他们的偏好？Building Robotics开发了一个名为Comfy的应用程序，使居住者能够立即影响他们的热环境，并使用此反馈来优化满意度和HVAC能耗。Comfy已经运行了三年多，每天都有成千上万的用户使用这种新型的HVAC控制。使用数据显示了居住者日常热偏好的重要模式。这些数据可用于更好地了解恒温器温度设定点与居住者偏好之间的关系，改善恒温器控制，并为节能HVAC操作提供信息。
Occupants of commercial buildings tend to have restricted means of affecting HVAC operation. As a result, opportunities for energy efficient behavior change are limited to provisions over which the occupant has direct control. An occupant can turn off lights at night, unplug electronics, and other similar actions that they can adopt in their daily routine. However, these behavior changes fail to address HVAC energy consumption, a class of energy consumption that could benefit greatly from occupant feedback. A system for gathering thermal comfort preferences would create opportunities for energy savings, but how can occupants be incentivized to express their preferences? Building Robotics has developed an application called Comfy that gives occupants the ability to immediately influence their thermal environment, and uses this feedback to optimize satisfaction and HVAC energy consumption. Comfy has been in operation for over three years and has thousands of users engaging with this new style of HVAC control every day. The usage data show important patterns in the daily thermal preferences of occupants. This data can be used to better understand the relationship between thermostat temperature setpoints and occupant preferences, improve thermostatic control, and inform energy-efficient HVAC operations.