Damage Identification Research For Continuous Girder Bridge Based On Support Reaction Influence Line

During the service period of bridge,owing to the coupling effects of various unfavorable factors,bridge structure generates damage accumulation and resistance attenuation inevitably.If the bridge damage can not be found and repaired in time,when the damage accumulated to a certain extent,it may lead to the occurrence of bridge collapse accident.For medium and small span bridges,compared with the arrangement of large-scale health monitoring system,monitoring the reaction influence line of one or several bridge bearings to identify damage has lower cost and easier operation.Therefore,for a large number of continuous girder Bridges,it has important theoretical value and practical significance to carry out damage identification research based on reaction force monitoring of bearings.This paper focuses on the topic of damage identification method based on support reaction influence line of continuous girder bridge,and carries out a detailed and indepth study.The main research contents include the following four parts:(1)damage sensitivity analysis of continuous beam support reaction influence line index;(2)damage location based on the curve fitting of the measured support reaction influence line;(3)damage quantification by fusion of support reaction influence line and machine learning algorithm;(4)verification of damage identification method for continuous box girder bridge.In the second chapter,the analytical expressions of the support reaction influence lines before and after the damage of two spans continuous beam are derived based on the force method and three kinds of damage indexes of continuous beam bridge are constructed based on the support reaction influence lines.Then,by setting the key parameters in the expression,the change rule of the three support reaction influence line indicators is analyzed,and the sensitivity of the damage indicators to the measured noise is further studied.The results show that the second derivative index of support reaction influence line change is the most sensitive to damage and can accurately identify the damage location,but it is also more sensitive to the test noise.In addition,the index of support reaction influence line change has strong anti-noise performance and good positive correlation with the damage degree,so it has the potential to quantify the damage degree.In the third chapter,the paper proposes to use the second derivative index of support reaction influence line change as the damage location indicator to identify the damage location of continuous beams,use the cubic smoothing spline function to fit the measured influence line change to solve the problem that the damage indicator is too sensitive to noise,and finally combine the D-S evidence theory to fuse the multisensor information to achieve the accurate damage location.The feasibility and accuracy of this method for continuous beam damage location identification are verified by multiple numerical calculations and laboratory aluminum beam experiments.In the fourth chapter,a damage quantification method based on support reaction influence line index and machine learning theory is proposed.Based on the correlation between the change index of support reaction influence line and the damage degree,principal component analysis algorithm and extreme learning machine algorithm in machine learning theory are introduced to achieve the damage quantitation finally.The numerical simulation of a two-span continuous beam and the experiment of an aluminum beam show that the method can accurately identify the damage degree of the continuous beam.In the fifth chapter,the paper establishes a finite element model based on the Xiamen BRT continuous box girder bridge,and further extends the damage identification method proposed in this paper from continuous girders to continuous box girder bridges.The effects of the lateral deflection of the test vehicle and the local damage to the support reaction influence line are analyzed in turn.Finally,the singleloss and multi-loss conditions are set to verify that the damage identification method proposed in this paper can also accurately identify the damage location and extent of continuous box girder bridges.