Structural Safety Evaluation Of Bridge Based On Time Series Analysis

Influenced by various uncertain factors in the long-time operation of the bridgestructure, the hazardous events, such as the overrun structure, extended service,fatigue and corrosion and so on, will lead to serious safety hazards happened in thebridge structure, which may bring huge loss of life and property. Therefore, the abilityof finding the damage and then giving warning prompts timely are demanded in thesecurity assessment system. However, as the core of the bridge health monitoringsystem, the security assessment is developed on the basis of the security method ofthe accurate structural model. In practically, the structure parameters vary constantlyin the service of the structure, which makes that it is difficult to obtain satisfy resultsfor establishing the security assessment method of the accurate structure model.The magnanimity timing data, which can reflect the feature information of thestructure injury, is collected by means of the bridge health monitoring system, diggingout the evolution reflecting the structure system performance. The timing analysismethod can not only overcome the difficulty of establishing the accurate structuralmodel, but also possess versatility. However, its research is still in the primary step,and it is great practical significant for its deep research.Based on this, from the point of departure time series analysis, the securityassessment method and implementation technology for the bridge structure areexplored and researched in this paper. The main content is summarized as following:(1)The importance for the security assessment of the bridge structure is discussedand then the existed security assessment methods are summarized at the same time,analyzing its shortcomings, which highlights the importance of the method of timesequence analysis.(2)The timing analysis and the grey prediction model are introduced into theanalysis of the bridge monitoring data, and then the data processing method based onthe algorithm of ARMA-GM is proposed; and then the experiment results show thefeasibility of the ARMA-GM in the field of the prediction of bridge future state.(3) A security assessment method for the bridge structure is proposed on thebasis of ARMA, which combining the control limits of the control chart with thereliability indexes of the bridge structure and setting the primary warning alarm limitsand safety limits. The security condition for the bridge structure is judged through the distribution of the observations statistics in the two-degree control limits, resultingthat the control limits for the control chart is closely linked to the security designspecifications for the bridge structure.(4) The practical application for the proposed method is developed and applied inthe structural safety assessment in Masangxi Yangtze River Bridge in Chongqing as areal example.The results indicate that the monitoring data in the bridge health monitoringsystem can be processed through timing model and grey model, obtaining highforecast accuracy of the state parameters for the bridge structure. Based on this, thegrey AR combination model proposed in this paper can predict the trend of the bridgestructure in nearly time and can give warning in advance on condition that the step ofthe prediction is same. The security assessment indexes, determining by the method ofsecurity assessment based on ARMA control chart, can identify the structural damageeffectively, overcoming the shortcomings of vain alarms and alarm omission.