Urban Fire Risk Assessment Indexes

With the continuous improvement of urbanization in China,urban development has entered an important transition period.The original fire safety problems still exist,and the continuous emergence of super high-rise buildings,super-construction single buildings,large-scale underground buildings,and new logistics parks have brought new challenges to fire safety work.Studying the city's fire risk assessment indicators can provide scientific and technical means for the overall decision-making of the city's overall fire planning,emergency force layout and fortification forces,and has great practical significance for urban security and social stability.This article summarizes the fire risk assessment indicators based on previous research findings of fire risk assessment,and completes the comprehensive urban regional fire risk assessment index system.Based on the detailed data collection,the random forest algorithm is used to analyze the fire risk in urban areas.The contribution rate of the index to the fire occurrence rate,and the use of Pearson correlation analysis of the internal relationship between the indicators,based on the results of the random forest algorithm and correlation analysis,mainly obtained the following results:(1)By summarizing the research results at home and abroad,a more comprehensive urban regional fire risk assessment index system has been formulated,including 4 first-level evaluation indicators,11 second-level evaluation indicators,and 36-level evaluation indicators.(2)A total of 19 indicators with a cumulative contribution rate of 95% for the fire incidence rate were obtained by the random forest algorithm.The order of importance was: the average annual fire safety complaint rate,population density,and density of chemical key target units.High-rise building ratio,public fire safety satisfaction rate,fire station protection area,5-minute arrival rate,proportion of industrial buildings,ratio of underground buildings,per capita GDP,public gathering place buildings,proportion of population with college education,number of fire drills,ratio of migrants,million Number of firefighters,fire hydrant integrity,storage buildings,grade 1 or 2 fire-resistant buildings,and personnel filing rates;(3)A total of 12 indicators that account for 95% cumulative contribution rate of fire damage rate were obtained by the random forest algorithm.The order of importance was: storage building,micro fire station setting ratio,ratio of migrants,and number of firefighters,5 minutes arrival rate,proportion of the population with college or above,fire protection station protection area,underground building,fuel refueling station density,annual average firefighting complaint rate for people,fire-fighting unit density,public gathering place construction;(4)Through correlation analysis,it is concluded that the indicators with low importance selected by the random forest are basically not related to the fire occurrence rate or the fire loss rate;the annual average firefighting complaint rate of the ten thousand people has a significant positive correlation with the fire occurrence rate,and the proportion of storage buildings There is a significant positive correlation with the fire loss rate,which verifies the rationality of the order of importance of random forests and the rationality of index selection.