Bridge Construction Safety Monitoring For Intelligent Construction Site

Traditional bridge construction safety monitoring methods have problems such as wasted human resources and insufficient monitoring efficiency,and the use of information technology to empower it has become an inevitable trend.In order to better ensure the safety during bridge construction,relevant work has been carried out in this article that focuses on the two safety hazards,namely,failure to wear helmets and bridge stress abnormalities,which are easy to occur during bridge construction.Firstly,two intelligent helmet wearing detection networks are designed based on deep learning,then bridge stress monitoring methods are designed based on technologies such as Internet of Things,and finally a bridge construction safety intelligent monitoring system is designed based on technologies such as BIM lightweight and WEB.The specific work of this paper is as follows:(1)A helmet wearing detection method is desgined.In this paper,two helmet wearing detection network are designed on the basis of YOLOv4.One is the lightweight network,namely,YOLO-Ghost-Bi FPNs3 whose calculated amount is only 5.8% of YOLOv4.It is used in most scenarios where the monitoring distance is average and using it can effectively reduce the power consumption of computing devices.the design idea of this network is as follows: firstly,the Ghost module is used to reconstruct the network of YOLOv4 to reduce the amount of computation,then the network is cropped to remove the redundant parameters,and finally the Bi FPNs3 structure is designed and used as the neck network to improve the inference speed.The other is the high accuracy network,namely,YOLO-Res2C-H4 whose detection accuracy is 1.8 percentage points higher than YOLOv4.It is used in a few scenes with long monitoring distance and can effectively identify small targets.the design idea of this network is as follows: firstly,the Res2 C structure is designed and used to reconstruct the network of YOLOv4 to improve the network’s ability to mix receptive fields,then the network is croped to reduce the amount of computation,and finally a small target detection layer is added to improve the detection accuracy of the network.(2)A bridge stress monitoring system is desgined.In this paper,three types of stress sensors,namely,steel stress,concrete stress,and soil stress,are used to detect the stress at different locations of the bridge,and the stress data are obtained through signal acquisition instruments,serial servers,and industrial control computers.After the stress data is obtained,on the one hand,the industrial computer submits the abnormal stress data to the server for early warning through the WEB software;On the other hand,the industrial computer first uses wavelet threshold denoising algorithm to remove the noise from the stress data,and then solves the envelope which assist stress trend analysis to give early warning before the stress is abnormal.The process of wavelet threshold denoising is as follows: firstly,wavelet decomposition is performed,then the threshold function and threshold shrinkage function are designed,and the wavelet coefficients corresponding to the noise are filtered out by using the both functions.Finally the wavelet coefficients are reconstructed to obtain the denoised data.(3)The WEB service software of bridge construction safety monitoring is developed,and the algorithm of helmet wearing monitoring and the program of bridge stress monitoring are integrated and realized.Firstly,the model designed by Revit software is lightweight,and then the bridge stress monitoring,safety wearing monitoring and lightweight BIM are integrated into an intelligent monitoring system.The lightweight BIM model can provide the3 D visualization platform for safety monitoring data,and users can quickly obtain the location of stress anomalies and the location of people without helmets with the help of the3 D model.The BIM lightweight model can also be used for guiding bridge construction,assisting project management,and coordinating the work of all parties.The system can be used in mobile phones,users can remotely manage bridge construction through mobile phones,which greatly improves the informatization level of construction safety management.