【正文】 外文文獻： Estimates of the Operational Reliability of Fire Protection Systems For the past three years， the National Institute of Standards and Technology (NIST) has been working to develop a new encryption standard to keep government information secure． The anization is in the final stages of an open process of selecting one or more algorithms， or datascrambling formulas， for the new Advanced Encryption Standard (AES) and plans to make adecision by late summer or early fall． The standard is slated to go into effect next year． Background Fire protection strategies are designed and installed to perform specific functions. For example, a fire sprinkler system is expected to control or extinguish fires: To acplish this, the system sprinklers must open, and the required amount of water to achieve control or extinguishment must be delivered to the fire location. A fire detection system is intended to provide sufficient early warning of a fireto permit occupant notification and escape, fire servicenotification, and in some cases activation of other fire protection features (., special extinguishing systems, smoke management systems). Both system activation (detection) and notification (alarm) must occur to achieve early warning. Construction partmentation is generally designed to limit the extent of fire spread as well as to maintain the building’s structural integrity as well as tenability along escape routes for some specified period of time. In order to acplish this, the construction features must be fire “rated” (based on standard tests) and the integrity of the features maintained. The reliability of individual fire protection strategies such as detection, automatic suppression, and construction partmentation is important input to detailed engineering analyses associated with performance based design. In the context of safety systems, there are several elements of reliability, including both operational andperfornzance reliability. Operational reliability provides a measure of the probability that a fire protection system will operate as intended when needed. Performance reliability is a measure of the adequacy of the feature to successfully perform its intended hnction under specific fire exposure conditions. The former is a measure of ponent or system operability while the latter is a measure of the adequacy of the system design. The scope of this study was limited to evaluation of operational reliability due primarily to the form of the reported data in the literature. In addition to this distinction between operational and performance reliability, the scope focused on unconditional estimates of reliability and failure estimates in terms offaildangerous outes. A discussion of these terms is provided later in the paper. Scop This paper provides a review of reported operational reliability and performance estimates for (1) fire detection, (2) automatic suppression, and to a limited extent (3) construction partmentation. In general, the reported estimates for fire detection are largely for smoke detectiodfire alarm systems。 automatic sprinklers prise most of the data for automatic suppression, and partmentation includes partment fire resistance and enclosure integrity. It should be noted that in some cases the literature did not delineate beyond the general categories of “fire detection” or “automatic suppression,” requiring assumptions regarding the specific type of fire protection system. Several studies reported estimates of reliability for both fire detection and automatic sprinkler system strategies. However, very little information was found detailing reliability estimates for passive fire protection strategies such as partmentation. A limited statistical based analysis was performed to provide generalized information on the ranges of such estimates and related uncertainties. This latter effort was limited to evaluation of reported data on detection and data were identified on partmentation reliability to be included. This paper addresses elements of reliability as they relate to fire safety systems. The literature search that was performed for this analysis is reviewed and important findings and data summarized. The data found in the literature that were applicable to sprinkler and smoke detection systems reliability were analyzed, with descriptive estimates of the mean values and 95 percent confidence intervals for the operational reliability of these in situ systems reported. Elements of reliability analysis There is considerabIe variation in reliability data and associated anaIysesre