The Safety Assessment Of High-rise Building Construction Based On The BP Neural Network

Urbanization is advancing and science and technology is progressing. Construction industry is developing very quickly. Especially tall buildings have sprung up. All over the country are constructing kinds of skyscrapers. China’s skyline high is constantly refreshed. A lot of high-rise building safety accidents happen every year. Every year thousands of one thousand people are killed. Thus it brought huge economic losses. Although the rising level of construction safety management and technology and the development of the legal system in china, the absolute number of safety accidents each year is much more than Europe and the United States and other developed countries. So on the research of security problems, the situation is still not optimistic.This article studies and analyzed the characteristics of the high-rise building construction safety production. On the summary of a large number of high-rise building construction accident case, the accident types of high-rise building construction in recent years in our country are Summarized. They were falling, hitting, collapse, mechanical damage, lifting damage and electric shock accidents such as type. Every kind of damage types are analyzed in detailed. The cause of the high-rise building construction safety accident is Summarized. The index evaluation system of high-rise building construction site safety is established on this basis.The safety evaluation of safety evaluation is nonlinear. BP neural network has a lot of advantages, such as strong nonlinear approximation, the training study, massively parallel processing, fuzzy reasoning, good fault tolerance, organization and so on. These characteristics of artificial neural network will be a kind of innovation of high-rise building construction safety evaluation. Its superiority is to conquer the defects of the old safety evaluation method, which brings a the new evaluation model of data processing. It evaluate the safety of the high-rise building construction accurately and quickly.The training samples are necessary for BP neural network. Through the research, this paper decides to use expert scoring and AHP method to obtain more objective, accurate and real sample data. This is very important for subsequent research work. MATLAB is a professional software using neural network. Finally the trained neural network is used into practice. In this paper, a project of Dalian is selected for safety evaluation, which is under construction, and the guidelines are presented.