AI-based Automation for Commercial Office HVAC: A Verdigris Case Study

• 传感器 - 环境传感器
• 传感器 - 温度传感器

• 建筑物
• 水泥

• 设施管理
• 维护

• 楼宇自动化与控制
• 占用率监控

• 系统集成

今天的建筑物运行时间更长，支持更广泛的最终用途，并支持更高水平的经济生产力，从而产生更小的错误余地。但即使是最先进的建筑和环境控制系统也未能实现设施和运营管理。我们的建筑物效率低下，使用它们的人服务不足。

为了满足乘员的舒适度并保持成本和能源效率，需要一种动态的人工智能辅助方法。

世界500强商业写字楼

财富 500 强是美国 500 家最大公司的年度名单，按其各自财政年度的总收入排名。

Verdigris 开发了一种解决方案，可在保持居住舒适度的同时自动减少 HVAC 能源使用。

Verdigris 传感器每秒监测建筑物的能源使用数千次。为了创建预测，人工智能添加了当地天气、公用事业定价、建筑管理系统 (BMS) 数据和其他可用数据集。如果数据可能不明确可用，Verdigris AI 可以推断出来。在这个模拟中，Verdigris 的人工智能通过观察安全徽章模式来了解建筑物的占用情况。

本报告模拟了财富 500 强运营建筑的自动且持续优化的暖通空调 (HVAC) 系统的结果。

• The AI-based solution developed by Verdigris was able to significantly optimize the HVAC system of the Fortune 500 company's building. The solution was run with three parameter selections: optimizing for temperature stability, optimizing for occupant responsiveness, and a balance of both. All models indicated the potential to optimize the current HVAC systems. For instance, HVAC systems were found to operate at levels more than necessary to maintain a comfortable occupant environment during the weekends when occupancy is minimal. Under Verdigris AI-optimized HVAC operations, compliance with comfort objectives increased to 100% for both the stable temperature and balance objective scenarios. In an occupant responsiveness optimization scenario, there were some time intervals during which the temperature may fall outside comfort zone guidelines, but all of those occurred during times the AI engine forecasts little to no occupancy.

• Persistent automated HVAC energy savings up to 18.7%
• Persistent automated HVAC cost savings 22.7-33.7%, depending on optimization criteria
• Increase from 4.5% of “occupied hours” within ASHRAE 55 standard to a persistent automated 100% of time within optimal productivity performance