【正文】 unction may be incorporated within the former type of system. BAS is used herein to refer to a management system that includes energy management features.BASs typically have the capacity to record historical operating data, referred to as “trend” data. BAS trending is a mon base system feature [3]. Typical trend data include temperature, humidity, valve and damper positions, on/off control signals, and air flow rates。 rarely are thermofluid flow rates and subhourly electric demand available. The number of BAS sensors in buildings is very large, so manual missioning is laborintensive and, therefore, automatic tools using trend data to aid in the missioning process are preferred [4]. The puting power and wireless capacity of BASs has made them economical and feasible for widespread adoption in buildings。 however, trend data are rarely used effectively to maintain optimal energy performance. Currently, BASs perform poorly when diagnosing HVAC faults, yet are essential to provide data for missioning [5].There is a need for automatic software tools to retrieve HVAC trend data to understand operation and performance, analyze variables of interest or performance indices required for ongoing missioning, and automatically update the input files of energy analysis programs to assist in the calibration process.This paper continues with a literature review on use of trend data in ongoing missioning and calibrated simulation. That section will be followed by a presentation of the HVAC system in a case study building, and the trend data available from the BAS. Subsequently, a proofof concept prototype, called the Automatic Assisted Calibration Tool (AACT), will be presented, with examples of how this tool can assist in ongoing missioning and generate inputs for building simulation. The paper finishes with a discussion of the AACT and its current limitations.2. Literature review. Use of trend data in ongoing missioningUse of trend data in ongoing missioning was researched and developed over the past 15 years with a focus on fault detection and diagnosis (FDD) [6–12]. These FDD studies found that BASs were practi cal and costeffective in providing the necessary data to conduct FDD research.Friedman and Piette [3] pared manual and automated FDD tools using HVAC trend data. They noted that one of the advantages of these tools was their capacity to reduce data management and analysis time to obtain information from trend data. Piette et al. [13] developed a prototype system using dedicated sensors, data acquisition software and hardware, and data visualization software including a webbased remote system. Their system included sensors additional to those avail able in mon BASs. They used their system to identify control prob lems and faults specific to HVAC systems. Seidl [14] documented his experience with troubleshooting control systems using trend data. He calculated how much a trend point deviated from its set point or from a benchmark to identify faulty trend points.Wang and Wang [15] presented an automatic missioning soft ware tool that used a large trend database to verify and diagnose the sensors used by the refrigeration systems during normal operation and during the missioning stage. They found the software provided a convenient and reliable means for the engineers to check and diagnose the BMS measurement devices.Katipamula and Brambley [1] stated that trend data are rarely used in industry because FDD is rarely built into BASs and there is a lack of infrastructure to gather data from existing BASs for addon applications. Increasing the automation of FDD methods with BASs was strongly advocated [4,8,16].Xiao et al. [17] presented an online diagnostic tool to monitor the condition of sensors used in an airhandling unit, applying FDD to trend data. Schein [18] developed an information system for a mediumsized office building BAS。 the information system collected design, operation and maintenance data about mechanical, electrical, and plumbing systems. The system was designed to extract trend data for other applications such as energy simulations and building missioning.Choiniere [19] presented Diagnostic Agent for Building Operation (DABO) developed by Natural Resources Canada that could be integrated into existing BASs to aid in automating ongoing missioning. DABO automatically analyzed trend data to identify faults using a rulebased reasoning module, provided suggestions to improve performance and generated energy and fort profiles.Wang et al. [20] presented an online fault diagnosis tool for VAV terminals that used monly available sensor and control signal trend data.The accuracy of trend data has often been questioned. The use of dedicated and calibrated sensors was expected to provide higher quality data than those installed in BASs. Haves et al. [21] listed the main issues with sensors: improper positioning, inadequate calibration during missioning, and drift during operation. Torcellini et al. [22] re mended the use of dedicated systems for subutility monitoring but this presents challenges in terms of additional cost and plexity.. Use of trend data in calibrated simulationUse of trend data in calibration appears to have first been reported by Carling et al. [23]. The trend data they used included BAS set points and control signals, submetered electricity, air temperatures in the AHU and zones, and slab temperatures. They identified multiple faults that might have been missed without trend data.Monfet et al. [24] used trend data to calibrate an energy model for AHU supply air flow rates and supply and return air temperatures, and the supply and return temperatures for hot and chilled glycol. They noted that trend data revealed operational faults and improved their initial input file.Pang et al. [25] used trend data to create inputs for realtime use of an already calibrated model.Gestwick and Love [26] used trend data to estimate lighting,