Study On Construction Monitoring Of Concrete Box Girder Bridge Based On BP Neural Network

In recent years,bridge construction in China is developing rapidly.As bridge span is increasing,bridge structure is also innovating,and bridge system is expanding.As a result,all kinds of accidents in bridge engineering are also becoming more and more serious,which are mainly manifested in two aspects: the increase in the possibility of accidents and the increase of the severity of the engineering accidents consequences.In order to ensure the safety of bridge construction process and meet the specification requirements of the linear and internal forces when the bridge is completed,a safe and reliable bridge monitoring system has great significance to the engineering application.Based on the existing bridge monitoring methods and the construction characteristics of prestressed concrete continuous box girder bridge,aiming to study the rationality of the monitoring methods based on BP neural network,this study gives a detailed theoretical derivation and verifies with practical engineering example.The study mainly carries out research in the following aspects:(1)Introduce the development and research status of bridge monitoring,summarize the contents and basic methods of bridge construction monitoring,focuse on the construction error adjustment methods of bridge monitoring,explore the characteristics of each method,explain the principle of action,list the advantages and disadvantages,theoretically expound the practicability superiority of the neural network method in the construction monitoring of prestressed concrete continuous box girder bridges.(2)Elaborate the concept,development and characteristics of artificial neural network and the mechanism of BP neural network,summarize the importance and feasibility of using BP neural network.Deeply study the mathematical principle of BP artificial neural network algorithm,clarify the learning mechanism of BP algorithm,derive the related formulas and use the MATLAB software platform to build BP neural network.(3)Summarize the problems existing in the traditional BP algorithm,and put forward improvement measures which are combined with the need of the practical project.Aiming at the defects of BP neural network structure,such as low learning efficiency and slow convergence speed,some improvement measures are proposed.Then achieve the improved BP network program expression through the MATLAB software platform,check and verify the network according to some cases to get prepared for practical engineering application.(4)With the construction background of the concrete continuous box girder bridge in Beijing new airport line program,monitor the stress of main structural components during the construction,use the MIDAS/Civil software for structural simulation analysis.The results are used as the learning samples of BP neural network for construction feedback and construction forecast.Comparing the results of BP neural network with the measured values,the study concludes that the improved BP neural network works well in ensuring the safety of the bridge construction process,and proves the effectiveness of improved BP neural network.