Research And Application Of Intelligent Troubleshooting Model Of Hidden Dangers In Construction Site Based On YOLOX

In recent years,although the accident rate in the construction industry has been decreasing,the safety production situation is still grim.The investigation of hidden dangers in construction is one of the important means to restrain the occurrence of construction accidents.The traditional hidden danger investigation mainly adopts manual methods,which has the disadvantages of being time-consuming,labor-intensive,and inefficient.The rapid development of computer vision technology,it has provided effective support for construction safety management.Among them,the target detection technology does not require construction workers to wear additional complex sensing equipment,which has less impact on construction site operations and can process on-site monitoring data in real-time to achieve rapid detection of hidden dangers on the construction site.However,the research and application of the current target detection technology to construction safety management is less.This paper takes the hidden dangers of the construction site as the research object and builds an intelligent investigation model of the hidden dangers of the construction site.First of all,based on the literature review,analyze and summarize the types of hidden dangers in various construction engineering scenarios,and list the hidden dangers of each subsystem,including familiar temporary works,earthwork works,foundation pit works,concrete structure works,and steel structures.At the same time,it also analyzes the safety protection equipment and related hidden dangers involved in various projects.It provides an application basis for different scenarios for constructing an intelligent inspection model for hidden dangers in construction.Secondly,the YOLOX(You Only Look Once X)target detection algorithm is selected to build an intelligent troubleshooting model for hidden hazards,including the arrangement of acquisition equipment,data collection of hidden hazards,image data annotation,data set production,model training,model performance evaluation,etc.;and elimination,and select the hidden dangers suitable for the model in each engineering scenario for analysis.Then select the electric welding operation area of the steel structure project construction site for intelligent investigation and application of hidden dangers,and start the work with the goal of wearing safety helmets,welding masks,masks,and welding gloves.The results show that the intelligent inspection model for hidden dangers in the steel structure engineering scene of the building construction site has an average precision(m AP)of82.28% for the intelligent inspection of hidden dangers in the electric welding operation area,and the performance exceeds that of the other four mainstream target detection algorithms,and the analysis results are time-sensitive.Finally,on the basis of the hidden danger intelligent screening model,the paper determines the weight of each hidden danger in the area and formulates the hidden danger scoring rules through the analytic hierarchy process to realize automatic safety evaluation.Compared with the traditional on-site inspection by security personnel,the appeal method can realize real-time and rapid hidden danger investigation and obtain more comprehensive and objective results,and also reduce the work risk of security management personnel.Through the research on the hidden dangers of construction sites,the paper helps to identify,warn,control,and eliminate hidden dangers in construction sites on the one hand;Safety accidents occur on construction sites,and the level of safety management on construction sites should be improved.The paper has a picture 30,a table21,and a reference 92.