【正文】 . Technologically, intelligent building performs and arranges differently from a conventional one. First, intelligent building technologies are characterized by a hierarchical presentation of system’s integration. Building systems and structure integration, as pointed out by Bradshaw and Miller , are to provide the ‘qualities that create a productive and efficient environment such as functionality, security and safety。 and building integrity’. According to Carlini and Arkin and Paciuk, many intelligent buildings prise three levels of system integration which include: ? The top level which is dealt with the provision of various features of normal and emergency building operation as well as the munication management。 ? The bottom level which contains subsystems including heating, ventilation and airconditioning (HVAC) systems, lighting system, fire protection system, vertical transportation system security system and munication system. Moreover, intelligent building allows interaction and integration among building 16 subsystem services. System integration is the process of ‘connecting systems, devices and programs together in a mon architecture so as to share and exchange data’. Arkin and Paciuk suggested the key to the effective operation of intelligent building was not related to the sophistication of the building services systems, rather it was the integration among the various systems, between the system and the building structure. Examples of major intelligent building systems interaction include: ? Fire alarm system would be integrated with other building systems, such as HVAC, lighting and security through BAS. HVAC systems can be used to prevent the smoke from spreading by opening exhaust dampers and closing outdoor air intake dampers of the fire floor if there is a fire on one floor of building。 ? Fire alarm program would be interfaced with security to release specific locked doors under alarm conditions。 ? Facility management is integrated with BAS. Ivanovich reviewed current research in intelligent building technologies. Contemporary research efforts have been attempting to develop software with the use of automated diagnostic tools introducing neural works, fuzzy logic, as well as other software intensive, artificialintelligencebased technologies designed to detect problems, such as building services systems/ponents, sensors and control devices that are hard detected by humanbeings. The IEA BSC research program Annex 25 and Annex 34 , involving over 10 universities and research institutions from different, conducted extensive research on the methodology, strategy and application of fault detection and diagnosis in HVAC systems. Many researchers also paid efforts in developing intelligent control method to be used in modern building management system for improving and optimizing the energy and environmental performance of buildings. In addition, the application of wireless technologies with buildings or 17 working building systems is also a popular research area which has currently attracted attentions of many researchers and industry practitioners. Research in performance evaluation methodologies Apart from intelligent technologies research and development, there have been substantial amount of research devoting to evaluating intelligent building. Serafeimidis considered the evaluation process as a feedback mechanism aimed to facilitate learning, while considered the process of evaluation as ‘a(chǎn) series of activities incorporating understanding, measurement and assessment. It is either a conscious or tacit process which aims to establish the value of or the contribution made by a particular situation. . and can relate the determination of the worth of an object’. Building performance evaluation is a crucial procedure which offers feedback function on the performance of building materials and ponents for future improvement and reference (: cited in Ref.). Different authorities have tried to develop evaluation models to assess the performance of intelligent building . Early performance evaluation models were developed by Manning in 1965 and Markus et al. in and Schramm later (in 1997) improved their evaluation models and proposed an ‘integrative building performance evaluation framework’ to evaluate and review the stance in all six major phrases of building delivery and life cycle including planning, programming, design, construction, occupancy and recycling. Many similar studies have also been conducted attempting to measure the level of intelligence that a building exhibited and to set up criteria for selection of the best intelligent building .Preiser developed the ‘postoccupancy evaluation process model (POE)’ in order to determine the intelligence level of intelligent buildings. The POE process model is generally executed in three stages. First, to develop patible data collection instructions in the conceptual phase。 third, to carry out parative analysis of data collected and development of remendations and guidelines for the utilization of the datagathering instruments worldwide. Preiser and Schramm applied the POE process model to evaluate intelligent building in the crosscultural context and suggested that the POE model could ‘enhance building performance evaluation in 18 intelligent buildings especially in a longterm, continuing basis’ because the evaluation system allows the ‘tracking of performance of new hightech systems and their effects on building occupants as well as the effectiveness of these systems in