Online system optimization algorithm development for building energy management

随着对温室气体排放的担忧增加以及未来能源成本的不确定性，提高建筑能效的需求正在增加。对建筑系统的监控对于收集信息很有价值，但尽管在开发人工智能（AI）支持的控制系统方面进行了大量研究，但商业系统通常仅限于更简单的算法，需要大量时间来部署和定制这些系统以适应特定的建筑，并进行季节性调整。因此，使用先进的人工智能方法开发自动控制调谐以允许在线学习和季节性调整是非常可取的。该项目扩展了之前的ENGAGE项目，该项目展示了该方法在供暖系统中的价值，并将其复制到冷却季节优化中，以最大限度地减少能源消耗，同时保持热舒适条件。我们将以Engage资助下供热算法开发的先前进展为基础，在正在进行的现场研究中测试应用程序，为未来的改进提供信息，并纳入更积极的天气调整控制预测。一旦采暖季结束，我们将收集冷水机和建筑制冷系统数据，以建立多单元住宅制冷系统的预测负荷模型和控制点优化模型。这些算法将使用基于云的建筑监控系统进行开发和部署，并从资助的第4个月开始每月向行业合作伙伴提供更新的算法，以便在冷却系统期间进行现场测试和持续改进。通过与现有设施合作，算法将在试点项目中进行测试和部署，以确认实际建筑物的实际节能效果。

The need for improved energy efficiency in buildings is increasing as concerns over GHG emissions rise and uncertainty over future energy costs. Monitoring of building systems is valuable to gather information but despite substantial research to develop artificial intelligence (AI)-supported controls systems, commercial systems are typically limited tosimpler algorithms and require significant time to deploy and tailor these systems to a particular building, alongwith seasonal adjustment. Therefore, the development of automated control tuning using advanced AIapproaches to permit online learning and seasonal adjustment is extremely desirable.This project extends a previous ENGAGE grant that demonstrated the value of this approach for heating systems and replicate it for cooling season optimization, to minimize energy use while maintaining thermal comfort conditions. We will build on previous advances in heating algorithm development developed under the Engage grant and test the applications in ongoing field studies to inform future refinements and incorporate more aggressive weather-adjusted control predictions. Once the heating season is over, we will collect chiller and building cooling system data to develop both a predictive load model and set of control point optimization models for the cooling systems for Multi-Unit Residential Buildings. These algorithms will be developed and deployed using a cloud-based building monitoring and control system, with updated algorithms provided to the industry partner on a monthly basis starting in Month 4 of the grant to permit in-situ testing and ongoing refinement during the cooling system. By working with existing facilities, the algorithms will be tested and deployed on pilot projects to confirm actual energy savings in real buildings.