【正文】 analysis to assess the average probability and 95% confidence intervals are estimated based on data provided by HALL [1955]. The scope of the average reliability estimates ranging from 6888%. Figures are consistent with those provided by the standards with Toby Delphy. Institute reliability. However，the general scope of the 95% confidence interval 6690%.Table 3： Analysis of the reliability of smoke detectors [HALL，1955]The fire district to rely on various types of equipment such as： doors (including fixed equipment)，wall，floor / ceiling penetration holes，windows，fire shutter，smoke materials and buildings. When the fire district is considered to be the focus in the fire plan，in the literature，there is little data that the individual ponents of the operating role in the fire district. The single is mentioned for the building assessment and operational reliability WARRIGTON research and the Australian Fire Engineering Index. These assessments are based entirely on expert judgment. Therefore does not provide more indepth analysis.Automatic sprinkler systems analysisOn Table 2，the sprinkler system reliability analysis is to analyze according to the type of each live. It should be noted that only one source MILNE，1959] on the reliability of public buildings and residential housing estimates，and these early data do not provide the reliability of the data of modern residential sprinklers. The distribution histogram of Figure 1 lists the reliability estimates for each housing type. Average and 95% confidence interval limit is suitable for general residential (in the study does not distinguish between mercial buildings，the category of residential buildings and public buildings) and mercial buildings，and is applicable to the building (mercial，public buildings，residential) reliability assessment. These results are shown in Table 4.Figure 1： The automatic sprinkler system reliability assessment of a variety of housing typesCommercial buildings and public buildings，reliability assessment，the average of control in other residential type 95% confidence interval. Livable and public buildings in a single point estimate，an increase of some useful things and operational reliability，and also increased the capacity of the database. 18 estimated four separate categories. However，the point on residential and public buildings of the estimates should not be used alone to draw any conclusions.The reliability of mercial buildings，residential and construction is estimated to provide some useful information. Based on the analysis of the sprinkler system can make use of the data is the reliability of operation is estimated at more than 88%，if you do not consider mercial buildings，the reliability of the sprinkler system can reach more than 92%. However，the judge this particular sprinkler system with those mentioned in the assessment system similarity is very important. The reliability of the mercial buildings is in the range 8098%，9498% in the general construction. Fire detection system analysisThe data on estimates of the reliability of fire detection system is prehensive. This data spans a decade，and every year the Reliability Assessment Report (reflected in Table 3)，it is done for the housing of a variety of different uses. Where the analysis is based on the use of the house divided into several building level. After each use housing derived data and then calculate the reliability of each housing estimates. Figure 2 shows the reliability estimates of all smoke detectors on all residential types.Figure 2： The smoke detector for a variety of residential types of reliability estimatesAs shown in the histogram，the data have a twostate distribution. Therefore，in order to further study the two average a pletely different database，a variance (ANOVA) analysis. Variance to detect the reliability of the estimated mean and reliability of a given type of construction.The graphical representation of the variance in the form of least squares method is reflected in Figure 3.Variance affect the final result. Shown in Figure 3，three residential classification of the respective estimated average reliability of smoke detection from. Histogram contained in Figure 4 describes the estimated reliability of each housing type.Figure 3： The smoke detection system variance analysis of the reliability of a variety of housing typesThese housing types in the average reliability estimates and 95% confidence interval estimate for separate analysis. The results are listed in Table 5，for each type of results are significantly different. Confidence interval from the spray system reliability for a variety of residential types of estimated confidence intervals do not overlap. This may make more data for the analysis of the smoke detectors，are listed in Table 5，the reliability of various residential types of smoke detectors is estimated to determine the reason for the difference of nonrelevant data beyond the scope of this analysis.Summary and conclusionsThe above analysis can easily be applied to other fire operation of the system reliability assessment. However，it should be noted that the reliability of data in the literature is an important factor. It is noteworthy that the data in content and form of the tremendous changes in the study of reports and studies，which is part of the effort. This study provides a very preliminary attempt to describe the operational reliability of fire protection systems. The survey report requires a lot of data to change the database.The focus of this effort to obtain more specific data，populationwide system can provide a basis for the great improvement of the operational reliability of fire protection system，which led to the interest of design engineers，in addition the performance of this technology for engineers based on the rapid development of design concepts the analysis is also necessary.12