Research On The Traffic Load Identification Based On Fiber Bragg Grating Sensing Technique

In recent years,with the increasing in the proportion of highway in the communication and transportation,more attention has been paid to the load identification because of the ubiquity of the vehicle overload and the frequent traffic accidents.The load identification is an important branch of structure health monitoring(SHM),which can be perfected and supply more data to support the evaluation of bridge condition and traffic management.However,the existing methods for traffic load identification have some boundedness,so more study need to be performed.Based on the fiber Bragg Grating(FBG)sensing technique,a simple method is proposed for the traffic load identification,which can identify the parameters of vehicle in a high speed,including the number of axles,the space of axles,the space of wheels,velocity,transverse loading position and the gross weight.Combined with the numerical example and the test in the laaboratory,parameters of the vehicle are identified,then the validity and applicability of the method are verified.The main contents and conclusions are as follows:(1)Based on the FBG sensing technique,a simple method is proposed for the traffic load identification and some description of the project has been done,which can identify the number of axles,the space of axles,the space of wheels,velocity,transverse loading position and the gross weight.According to the peaks number and space of the FBG sensors' response,the number of axles,velocity and the space of axles can be identified.The strain curve can be fitted by the response of different locations,and the turning point of the strain curve is the loading position.Besides,the space of wheels can be computed.The response peaks of sensors can be extracted and it is necessary to make a correction when the vehicle deviates from the mid-span,then the unknown weight can be identified combing the strain load coefficient.(2)In the numerical example,after the strain of girder in every condition is extracted,the vehicle parameters can be obtained in terms of the proposed method.The results show that the number of axles,velocity and the space of axles can be identified accurately.Using the strain of proper location on girder,the error of the loading position and the space of wheels can be controlled within the range of 1.5%and 0.3%,respectively.A correction is made about the strain in the condition of deviating from the mid-span with the strain position coefficient.And the error of load identification can be controlled within the range of 1.5%.Adding Gaussian White Noise to the original strain data,there is small difference between the identification error of vehicle parameters and the error in the condition of non-interfering,which indicates that the proposed method is robust to noise.(3)The road model is cast in the laboratory and the vehicle load test is performed.The vehicle parameters can be identified according to the response in different location of girder.The results show that the number of axles can be obtained precisely.However,the error of velocity and the space of axles increases with the growing velocity and can be controlled within the range of 15%.The error of the loading location increases with the distance between loading location and the mid-span,which is distributed from 5%to 10%.The error about the space of wheels can be controlled within 5%.The relationship between the error of load identification and the velocity is monotone increasing approximately according to the proposed method.So,the compensation algorithm based on the damping of velocity is used to correct the results of load identification and the final error can be controlled within 15%.