Analysis Of The Causes Of Construction Accidents And Research On Accident Early Warning

With the vigorous development of construction industry,construction accidents happen frequently,and the situation of accident prevention is still severe.At present,the research on the causes of construction accidents is not detailed and in-depth,and the research on accident warning is in a booming stage,but the research has just started in the field of construction.Therefore,it is of great significance to comprehensively and deeply study the behavioral causes of construction accidents,explore the potential rules of accidents,and establish accident early warning models for improving safety ability and further preventing and reducing the occurrence of construction accidents.In this paper,behavior-based accident causation " 2-4" model theory is used to analyze the causes of five typical injury accidents in construction(namely height falling accident,object strike accident,mechanical injury accident,collapse accident,electric shock accident).That is,from the reasons of individual behavior(unsafe behavior of people,unsafe state of things,safety ability)and the reasons of organizational behavior(safety management system,safety culture)carried on a comprehensive and systematic analysis and research.The results show that "no safety education and training","not aware of the importance of safety training","lack of safety management system","not aware of the importance of safety" are the main reasons for the occurrence of five major injury accidents in construction,revealing the behavioral causes of construction accidents.On the basis of the above research,this paper deeply studies the behavioral causes of construction fire accidents,demonstrates the objective regularity of construction accidents,and provides data basis for the construction of accident warning model.Then,machine learning algorithm and naive Bayes theory model were used to comprehensively consider the unsafe behavior of people,unsafe state of objects,safety ability,safety management and safety culture,and 10 attribute variables including accident time,accident grade,accident area,unsafe action and the classification and value conditions of each variable were proposed,and an accident warning model was established.In order to improve the accuracy of the model prediction,the traditional Bayesian model was weighted and improved,while adding four attribute variables which were highly correlated with the occurrence of accidents:age of operators,health of operators,light conditions and weather conditions,and the effectiveness of the model prediction was verified through comparative analysis.According to the regularity of the causes of accident behavior and the effectiveness of the accident early warning model,the prevention measures of construction accidents are proposed,so as to provide a theoretical basis for efficient safety management in the field of construction.