Research On Fire Risk Assessment Of High-Rise Building Based On BP Neural Network

The high-rise buildings are integrations of economy, technology and culture in modern society. For their high stories, large spatial spans and complex functions, once fires occur, the extinguishments are difficult and may cause heavy losses and disastrous consequences. Therefore, the concerned standards in China have made clear the principle of the automatic fire-fighting in high-rise buildings, and its reliability of techniques in fighting fires have drawn more and more attentions in recent years.According to the problems existing in the fire risk assessment of high-rise building management and assessment now, the research project of the risk assessment based on BP neural network was proposed and studied, and the author's original idea and methods was given out. We will plot out three orders about fire risk of high-rising building. By accommodating parameter of choosing dormant floor and confirming numbers of dormant floor node in BP neural network and algorithmic ameliorating, we have constituted the model of fire risk assessment of high-rising building based on BP neural network. Because there have not been standard information to assess about the fire risk assessment of high-rising building, we will use fuzzy mathematics to confirm stylebook of BP neural network disciplinal, according to material example and some area frontal experience of high-rising buildings in some area, we have constituted educating stylebook of fire risk assessment about architectural configuration and fire load , the system of calling the police and putting out a fire, the system of evacuating safety and manage. Finally we have educated stylebook of fire risk assessment in high-building, and used the software of MATLAB to educe educating infer chart of each subsystem, and have constituted model of fire risk assessment about four subsystem in high-rising building, and have given an example to validate its validity. The final result indicates that BP neural network is an effective method and tool in the comprehensive safety assessment of construction installation sites.