The Early Warning Research Of High Fall In Construction Engineering Based On The BP Neural Network

The construction industry is one of the pillar industries of China's economy.In recent years,the industry has achieved rapid development.As far as the development of China's construction industry is concerned,both the safety management level and construction technology are gradually improving,the relevant legal system is continuously improved,and the supervision is constantly increasing.However,the number of safety production accidents occurring each year is still much higher than that of the developed countries in Europe and America.The construction industry is still facing serious problems in terms of safety management.Especially nowadays,all kinds of buildings built throughout the country constantly refresh the height of our skyline.With the continuous improvement of the height of the building,thousands of people lose their lives every year due to the accident of falling from the height of the construction project,which not only causes serious economic losses,but also endangers social stability.This paper mainly analyzes the high-level fall accidents that often occur in construction projects,including the high-rise fall during the process of frontier operations,suspended operations,hole operations,operating platform operations,climbing operations and cross-operations accident.Describe the characteristics of high work.Detailed analysis of each type of accident,summing up the cause of the fall accident in the high point of the construction project.On the basis of that,a safety early warning indicator system for falling accidents in high-rise buildings is established.The safety warning of falling from the height of the construction project is a nonlinear model.Therefore,BP neural network has many advantages that traditional methods can not compare with.Its superiority mainly lies in:it is more comprehensive and complete than the previous security early warning method,and also proposes a new mode for data processing,which can provide more effective security warning for the fall of construction engineering high place,which can bring accurate and fast results.Therefore,BP neural network is chosen as the research model.MATLAB software is a professional software using neural network model,which is used to train BP neural network.In this paper,the early warning index of the falling safety of the building engineering is taken as the input data of the neural network,and the six types of accidents in the high falling accident are used as the output data.Score the input of 20 sample items by expert scoring.The BP neural network was trained by applying the first 18 sets of sample data.The trained neural network model was verified by the other two sets of sample data,and the reliability of the neural network in the safety warning of falling from the height of the construction project was confirmed.Finally,the paper applies the trained BP neural network to the phase of a healthy bio-industry home project,and conducts a safety warning for the high-altitude accidents that may occur during the construction process,and analyzes the safety warning results.