توسعه روش دستی ارائه اطلاعات برای شناسایی اطلاعات مورد نیاز برای تجزیه و تحلیل عملکرد سیستم های HVAC

Heating, Ventilation and Air-Conditioning (HVAC) systems account for more than 15% of the total energy consumption in the US. In order to improve the energy efficiency of HVAC systems, researchers have developed hundreds of algorithms to automatically analyze their performance. However, the complex information, such as configurations of HVAC systems, layouts and materials of building elements and dynamic data from the control systems, required by these algorithms inhibits the process of deploying them in real-world facilities. To address this challenge, we envision a framework that automatically integrates the required information items and provides them to the performance analysis algorithms for HVAC systems. This paper presents an approach to identify and document the information requirements from the publications that describe these algorithms. We extend the Information Delivery Manual (IDM) approach so that the identified information requirements can be mapped to multiple information sources that use various formats and schemas. This paper presents the extensions to the IDM approach and the results of using it to identify information requirements for performance analysis algorithms of HVAC systems.

Heating, Ventilation and Air-Conditioning (HVAC) systems account for more than 15% of the total energy consumption in the US [1] and [2]. However, research studies show that about 10–40% of the energy used by HVAC systems is wasted due to degradation faults, such as biased or drifting sensors, malfunctioning controllers, stuck dampers and fouled coils [3], [4], [5] and [6]. Actively detecting faults requires continuously monitoring and analyzing the status of hardware and software components that are part of HVAC systems. However, due to the increasing complexity of HVAC systems, manually monitoring thousands of components in these systems is very challenging and impractical [7], [8], [9] and [10]. Hence, researchers have developed computer algorithms in the past 20 years to automatically analyze and improve the energy performance of HVAC systems. Examples of these algorithms include computer-aided fault detection and diagnosis (FDD), automated commissioning and optimized operating schedules [11], [12] and [13]. Laboratory experiments have been conducted to validate the energy saving capability of these performance analysis algorithms [14], [15] and [16]. However, very few of real-world facilities have deployed them. One primary reason identified by several researchers is that it is very difficult for system operators to manually collect the information required by these algorithms. For example, Jagpal [7] discussed seven barriers to automated fault detection for HVAC systems, of which four are related to the difficulties associated with collection of the needed information: (1) detailed design data are seldom available; (2) measured data are unavailable; (3) some variables cannot be measured directly; and (4) design intent is poorly specified [7]. Similarly, Luskay et al. [8] emphasized the importance and difficulties of collecting needed information for automated and continuous commissioning tools [8]. In a previous study, the authors have analyzed these performance analysis algorithms and found that they require a variety of information, such as the dynamic sensing measurements and control signals, the configuration and specification of the HVAC subsystems and components, and the property of the building elements [17]. These information items are created by engineers from different disciplines including architecture, mechanical engineering and structural engineering, and managed using a variety of dispersed documents and software tools such as design drawings, spreadsheets, diagrams and manuals [18]. Hence, it is very challenging for system operators to manually collect and integrate all the information required by performance analysis algorithms. To address the challenge of deploying performance analysis algorithms in different buildings and HVAC systems, the authors previously proposed a self-managing framework that integrates heterogeneous sources of information about the building and HVAC systems and automatically provides the required information to different algorithms [17]. In order to automatically provide the required information, this self-managing framework first needs an information repository that contains all information required by different algorithms. These algorithms are described by researchers in academic and scientific publications, such as journal articles, conference papers, theses and research reports [9], [10], [11], [12], [13], [14], [15] and [16]. Hence, these publications need to be explored and analyzed such that information requirements contained in each publication can be identified and documented. The objective of the study described in this paper is to investigate a formal approach to identify and collect a general set of information requirements of performance analysis algorithms for HVAC systems. The Information Delivery Manual (IDM) approach was developed by the buildingSMART alliance to identify the processes that are undertaken within a building project and the information required by each process [19]. Its objective is to support the information specified by the Industry Foundation Classes (IFC) schema [20], which provides a standard way to represent and manage the semantic-rich information of building projects. However, according to our review of performance analysis algorithms for HVAC systems, it was found that IFC schema only covers about 60% of the information required by these algorithms. Hence, in order to increase the coverage of information requirements found in the publications that we studied, this research extends the IDM approach to identify and document the information requirements represented in multiple data models. The following sections will introduce the extended IDM approach and the results of using it in this context.

The study described in this paper presents an extended IDM approach to identify and document the information requirements of performance analysis algorithms for HVAC systems. The extended approach makes four changes to the original IDM approach. First, the extended approach identifies the opportunity to use two text-based tuples to represent the process maps. By documenting the process maps using text, the processes and exchange requirements are collected and the time required to create the information given in BPMN diagrams is reduced and the contents of the process maps can be analyzed automatically. Second, the extended approach adds a new step to the original IDM approach so that the functional parts can be directly identified from exchange requirements. This new step enables the identification of information items that are not included in the IFC schema. The third and fourth changes enable the functional parts to be mapped to multiple information schemas and the MVDs to be generated to retrieve information from multiple data models. This study then used the extended approach to map 91.4% of the functional parts identified from the performance analysis algorithms to three data models: IFC, gbXML and EnergyPlus. The results show that because the extended IDM approach is able to identify and document information requirements from different data models, the coverage of information requirements is improved by bringing multiple data models together. The results also indicate three challenges that are associated with using existing data models to represent information required by the performance analysis algorithms. First, there are information items that are not covered by existing data models. Second, the heterogeneous schemas and formats of different data models bring challenges to develop an approach that can generally integrate various data models. Third, because there are overlapped information items amongst different data models, integration of the data models also requires consistency verification of the values of overlapped information items. In order to develop the self-managing framework that addresses these challenges, the next steps of our research will investigate the approaches that extend data models, integrate heterogeneous data models, and verify consistency among information items from multiple data models.