Study On Safety Assessment Method Of Suspension Bridge Based On Health Monitoring Data Analysis Of The Qingcaobei Yangtze River Bridge

The development of bridge health monitoring and evaluation system is becoming more and more mature,which is widely used in the health monitoring of all kinds of bridge operation period,because the health monitoring system has real-time,visibility,long-term and stability.Although the use of health monitoring system provides a convenient operation space for the health monitoring of bridges,however,because of the large amount of data accumulated by the long-term monitoring can not be used reasonably,so that the importance of health monitoring is lost,so how to analyze these massive health monitoring data and make a reasonable assessment of the bridge,It has become a difficult point in the bridge health monitoring and evaluation system at present.Based on the research background of the health monitoring and evaluation system of the Qing caobei Yangtze River Bridge,this paper studies and analyzes the health monitoring data of the bridge of Qing caobei Yangtze River Bridge and establishes a safety assessment method for the suspension bridge under large data.The current mainstream bridge safety assessment method is comprehensive assessment method and analytic hierarchy process,for the comprehensive evaluation method,its essence is the optimized weighted average algorithm,which is similar to the analytic hierarchy process,and the determination of weights and the constitutive relation of data between the bridges can not be determined.Such a rating will often be authorized to change the value of the effect,will result in the assessment and the actual project does not match.In this paper,the data of the health monitoring and evaluation system of the grass-back bridge is analyzed and the safety status of the suspension bridge is evaluated with the neural network model.The main tasks are as follows:（1）Based on the finite element model of the Qing caobei Yangtze River Bridge established by Midas/Civil,the stress extremes of each node are obtained by the load influence line principle and different load combinations,and the monitoring thresholds of the Qing caobei Yangtze River Bridge are determined.Then taking the Qing caobei Yangtze River Bridge for one year of health monitoring data analysis The data changes of the main monitoring points were analyzed,at the same time,the total stress difference of the cross-section of the main girder and the tower is stripped by the stress of the section at different time of a continuous period,and the main girder of the Qing caobei Yangtze River Bridge is carried out on the other random live load except the constant load and traffic live load.The influence of stress day on the section of main tower is studied,it provides a reference for similar bridge research.（2）After comparing and analyzing the advantages and disadvantages of fuzzy neural network and BP Neural network,this paper establishes a fuzzy neural network structure based on T-S fuzzy neural network structure,adding the learning algorithm of BP Neural Network and the related parameter function.At the same time,the constitutive relation of suspension bridge and the mechanical boundary condition of main structure are combined.According to the input and output samples from the dynamic design and adjust the fuzzy system design parameters,the realization of the fuzzy system self-learning and adaptive function,so as to establish the optimized fuzzy neural network model of safety evaluation of the Qing caobei Yangtze River Bridge.（3）Based on the health monitoring data of the Qing caobei Yangtze River Bridge,this paper presents the methods and analytic hierarchy process and object oriented method^[]to evaluate the safety status of the Qing caobei Yangtze River Bridge,compares and analyses the advantages and disadvantages of the three methods,and sums up the advantage of the method presented in this paper.After comparing and analyzing the advantages and disadvantages of fuzzy neural network and BP Neural network,this paper establishes a fuzzy neural network structure based on T-S fuzzy neural network structure,adding the learning algorithm of BP Neural Network and the related parameter function.At the same time,the constitutive relation of suspension bridge and the mechanical boundary condition of main structure are combined.According to the input and output samples from the dynamic design and adjust the fuzzy system design parameters,the realization of the fuzzy system self-learning and adaptive function,so as to establish the optimized fuzzy neural network model of safety evaluation of the Qing caobei Yangtze River Bridge.