综合供暖、通风和空调（HVAC）系统是决定整个建筑能耗的最重要组成部分之一。对于商业建筑，尤其是办公楼和学校，热负荷和冷负荷在很大程度上取决于职业行为模式，如居住密度及其活动。因此，如果暖通空调系统能够对动态占用情况做出响应，那么就有很大的潜力来降低能耗。然而，目前大多数现有暖通空调系统都无法根据乘客的动态状况调整供气速率。 由于这种效率低下，暖通空调的大部分能源消耗都被浪费了，尤其是当有条件的空间无人或占用不足时（占用人数少于预期设计）。解决这一低效问题的方法是基于动态乘客特征控制HVAC系统。基于此，本研究提供了一个基于实时视觉的居住模式识别系统，用于居住计数和活动级别分类。研究分为两部分。第一部分是使用基于深度学习的开放源代码库进行实时占用计数，并使用静态RGB摄像头使用背景减法进行活动级别分类。 第二部分利用DOE参考办公楼模型，采用动态设定点控制和常规HVAC控制，以确定潜在的节能和热舒适性。研究结果表明，基于视觉的系统可以在没有太多遮挡的情况下实时检测乘客并对活动水平进行分类，准确率在90%左右。此外，动态设定点控制策略确实可以带来能量节省和热舒适性改善。引文:2018年冬季会议，伊利诺伊州芝加哥，会议论文

An integrated heating, ventilation and air-conditioning (HVAC) system is one of the most important components to determine the energy consumption ofthe entire building. For commercial buildings, particularly office buildings and schools, the heating and cooling loads are largely dependent on the occupantbehavioral patterns such as occupant density and their activities. Therefore, if HVAC system can respond to dynamic occupancy profiles, there is a largepotential for reducing energy consumption. However, currently, most of existing HVAC systems are being operated without the ability to adjust supplyair rate in response to the dynamic profiles of occupants. Due to this inefficiency, much of the HVAC energy use is wasted, particularly when theconditioned spaces are unoccupied or under-occupied (fewer occupants than the intended design). The solution to this inefficiency is to control HVACsystem based on dynamic occupant profiles. Motivated by this, the research provided a real-time vision-based occupant pattern recognition system foroccupancy counting as well as activity level classification. The research was divided into two parts. The first part was to use an open source library basedon deep learning for real-time occupancy counting and background subtraction method for activity level classification with a static RGB camera. Thesecond part utilized a DOE reference office building model with dynamic set-point control and conventional HVAC control to identify the potentialenergy savings and thermal comfort. The research results revealed that the vision-based system can detect occupants and classify activity level in real timewith accuracy around 90% when there are not many occlusions. Additionally, the dynamic set-point control strategies indeed can bring about energysavings and thermal comfort improvements.