建筑系统，包括供暖、通风和空调（HVAC）系统、照明系统和安全系统是现代建筑的关键组成部分。这些系统旨在提供理想的建筑环境，确保建筑服务质量和安全。楼宇自动化系统（BAS）的实施促进了楼宇运营和维护（O&M），使楼宇运营商能够更好地监控和调节楼宇功能。现代BAS的重要功能之一是在建筑系统的行为与设计值不同时发出警报。 然而，BAS通常每天都会产生过多的警报。这些警报缺乏可采取行动的信息，这使得建筑运营商采取行动非常困难。本研究的目的是找出BAS报警功能效率低下的原因，并提出解决方案，帮助建筑运营商根据BAS报警做出更好的运维决策。这项研究分析了一所大学建筑群中的一个BAS。首先，对大楼的暖通空调系统进行了调查。其次，对设施经理和BAS现场工程师进行了访谈，以确定BAS的现有缺陷和未来用户的期望。 第三，构建数据挖掘框架，优化现有BAS报警管理功能。该框架的目标是帮助过滤无关紧要的警报，按其影响类别对警报进行分类（例如，设备相关、职业相关、关键设备操作），并根据定量影响对警报进行优先级排序。本文为宾夕法尼亚州匹兹堡卡内基梅隆大学校园的111754平方英尺建筑群提供了一个警报管理优化解决方案。该研究发现，良好的BAS警报管理功能应该是设计师、BAS工程师和建筑运营商之间的协作。 当前BAS报警管理的缺陷是由糟糕的规则定义和信息隔离造成的。本研究的框架可以有效过滤BAS中的琐碎警报，并为建筑运营商提供可操作的信息。引文:2018年冬季会议，伊利诺伊州芝加哥，会议论文

Building systems, including Heating, Ventilation and Air-conditioning (HVAC) systems, lighting systems, and security systems are key components of modern buildings. These systems are designed to deliver an ideal built environment, ensure the quality of building service and safety. The implementation of Building Automation System (BAS) has facilitated the building Operation and Maintenance (O&M), allowing building operators to better monitor and regulate building functions. One of the important abilities of modern BAS is raising alarms when building systems behave differently from design values. However, BAS usually generates an excessive amount of alarms every day. The lack of actionable information from those alarms makes it very challenging for building operators to take actions. The intent of this study is to find the reasons of the inefficiency of BAS alarm functions and to propose a solution which helps building operators make better O&M decisions based on the BAS alarms. This study analyzed a BAS in a university complex. First, the building’s HVAC systems were investigated. Second, interviews with facility managers and BAS field engineers were conducted to identify the existing deficiencies of BAS and future user expectations. Third, a data-mining framework was constructed to optimize current BAS alarm management function. The goal of this framework is to help filter out trivial alarms, categorize alarms by their impact category (e.g. equipment-related, occupantrelated, critical equipment operations), and prioritize the alarms based on the quantitative impacts. This paper presents an alarm management optimization solution for an 111754 sqft-complex on Carnegie Mellon University campus in Pittsburgh, PA. This study found that a good BAS alarm management function should be a collaborative effort among designers, BAS engineers, and building operators. Current deficiencies of BAS alarm management are caused by poor rule definitions and information isolations. The framework from this study could effectively filter out the trivial alarms from BAS, and provide actionable information for building operators.