Life-cycle Prediction of Bridge Temperature-induced Deflection Extremes Based on Long-term Monitoring Data
Deflection is one of the important parameters to test the health status of bridges.It is of great engineering significance to accurately predict the extreme value of deflection in the whole life cycle.Based on the long-term monitoring data of deflection and temperature field of a bridge,the time-varying monitoring law of deflection of the bridge is analyzed,and the monitoring correlation law between deflection and seasonal temperature is investigated.By studying the cumulative probability characteristics of bridge deflection and its optimal cumulative distribution function,a new method for bridge deflection lifetime prediction considering the influence of seasonal temperature and diurnal random characteristics is proposed.The results show that the deflection of bridge has obvious seasonal variation characteristics,and has a good correlation with temperature.Therefore,the influence of seasonal temperature should be fully considered when predicting the extreme deflection value.The maximum value of deflection in each day shows stationary stochastic characteristics,so the influence of stationary stochastic characteristics of deflection maximum should be fully considered in the prediction of deflection extreme value.The stationary stochastic characteristics of deflection maximum can be described by probabilistic and statistical features.Compared with Normal and Weibull distribution function,GEV distribution function can better describe the probability and statistical characteristics of the extreme deflection.The extremum prediction method of deflection in the whole life cycle is proposed.The deflection in the whole life cycle is caused by seasonal temperature and stationary random characteristics.The extreme deflection of bridge in the whole life cycle is 190.14 mm.The research results can provide reference for service safety assessment of bridge structures.
bridge structuredeflectionlife-cycleprobability statistical characteristicsextreme value prediction
万方数据
维普
