Beijing Fire Hazards And The Gray Prediction Model

When a fire happened, it might be caused by several factors which were complicated and interacting. Although these complex factors of the particular matter are random, the variation of them, however, submits to a statistical regulation when the investigation is based on the whole city's fire matters. In order to provide some scientific conclusions of fire precaution and damage decrease, this paper researches the objective process and the regulations of the Beijing's fire matters in a mathematical and statistical way, and makes a scientific prediction about the future trend and the potential level of Beijing's fire damage parameters. In this essay, status research, classification research and prediction of Beijing's fire damage are made and the countermeasures are given in the methods of one-sample T test, cluster analysis and grey predicted model. The basic data is collected since 1997 and analyzed by Excel,SPSS and MINITAB.There are five chapters in this article. Chapter 1 is the introduction, which introduces the domestic and foreign fire damage research complexion, presents the study's purpose and significance, reviews the related theories and empirical literatures of fire damage, and indicates the research frame and methods. In Chapter 2, the Beijing's fire damage status in 2006 is analyzed by contrasting against the data of 1997. Chapter 3 is cluster analysis, which divides Beijing into three different parts—Haidian district, Chaoyang district and others—by the level of the fire damage. In Chapter 4, fire prediction is given. Chapter 5 presents the research results and the countermeasures.The content of this paper can be summarized into four parts. First, a more reasonable frame is built to support the research on the damage of the fire, causes of the fire, geographical distribution and the trend of change. Second, cluster analysis is used to divide Beijing into different fire damage areas. Third, Beijing's fire damage GM(1,1) gray predicted model is build up by using the grey theory. For the four parameters--the quantity of the fire matter, the number of the death, the number of the casualty and the grade of direct economical expense—can mainly reflect the level of the fire's damage, their smooth is verified and their basic data is pretreated by using transformation of inverse hyperbolic sine function. These means improve the model's precision and provide an effective way of predicting the fire time sequence. Finally, fire precaution countermeasures are offered and the workable and reasonable suggestions are given to support Beijing's future damage reduce work.