Research On Construction Safety Accidents Forecasting And Countermeasures In China

In recent years construction industry of China is in a rapid economicdevelopment, the government attaches more and more great importance to theconstruction safety management. A series of safety laws and regulations have beenissued one after another, and the building security situation is gradually improving.But compared to the developed countries, the safety management and accidentforecasting of engineering construction in our country are in much lower level. As oneof the high-risk industries, construction industry still has a serious security situationwhich has hindered the health sustainable development of the building and theforward pace of the harmonious society. Therefore, scientifically studied theconstruction safety accidents change tendency will be able to provide reference forour country safety in construction to prepare reasonable countermeasures.Firstly this article analyses and summarizes the research situation of constructionsafety management from the macroscopic and microscopic perspectives, whichinclude the accident-causing theory, safety technology,safety assessment, securitylegal system and so on, then points out the big gap between home and broad and theimportant significance of the construction safety accident forecasting. Secondly thispaper expounds the accident forecast theory, current research content and methods.The construction accident forecast is very complicated, One reason is that the relatedindexes have characteristics such as nonlinearity, randomness and uncertainty. Theother reason is the imbalance and complexity of the Chinese construction safetysituation. Considering the above characteristics and the objective situations, the paperadopted the grey theory method and the BP neural network method to forecast andsimulate the number of the accident occur and death.This article statistic the macroscopic data about the national construction from1998-2009, then build GM(1,1) model and BP neural network model to predict thebuilding development situation combined with the Matlab software. Before makingBP neural network model, a correlation analysis among the indexes of constructiondevelopment and the safety accident is used to select indexes that the coefficient ismore than0.5. At last,12indexes are obtained to help the BP neural network forecast.Based on the comparison between the simulation results and the practical statistics, the result of BP neural network is more accurate than the other. The results show thatthe BP neural network forecasting model combined with the macroscopic indexes hadthe ideal forecast ability.On the basis of the research, from the point of view of safety managementsystem, the construction safety accident prevention and control measures are putforward from the aspect of security technology, safety education, safety managementand insurance. The countermeasures have important and active meaning to theconstruction safety planning. Finally, the article made a conclusion and pointed outthe existing defects and the expectation for further research.