Identification Of Moving Force With Bridge Weigh In Motion System Based On MFI Theory

Bridge Bridge weigh-in-motion(BWIM)systems are based on measured strain of moving traffics passing an instrumented bridge and to estimate the axle weights of moving vehicles.The traditional commercial BWIM system uses influence line of the instrumented bridge to calculate the static axle weight while ignoring the influence of its dynamic effect.The application of moving force identification(MFI)theory in BWIM system can effectively combine the structural dynamic information to identify the moving vehicle load.Based on the field tests of two simple-supported T-beam bridges with different widths,relevant researches are carried out in this paper.The first-order regularization technique is combined with dynamic programming method and applied to the two-dimensional MFI theory to identify the axle load of a moving vehicle.Meanwhile,Eigenvalue reduction technique is used to reduce the degree of freedom of the dynamic system equations and the L-curve method is used to obtain the optimal regularization parameters.Based on the response signals measured by simulation analysis and field experiments,the axle weight-time history curve of the moving vehicle is identified in combination with the MFI theory,which verifies the reliability and accuracy of the proposed moving load identification algorithm based on the MFI theory.Specific research contents are as follows:(1)The two-dimensional MFI theory algorithm combines with the first-order regularization technique and dynamic programming method is deduced in detail,and it is combined with the bridge weigh-in-motion system.(2)Based on the simulation analysis of different vehicle driving bridges,the applied MFI algorithm is verified by the axle weights recognition of various cases.(3)Based on the measured dynamic response of two simple-supported T-beam bridges,the established models are tested and the MFI theory is used to identify the axle weight-time history curve of the moving vehicle.(4)Based on simulation analysis of vehicles traveling across different lanes with different vehicle speeds,the algorithm proposed in this paper can effectively identify the moving vehicle loads in different lanes.At the same time,the recognition accuracy of vehicles with different vehicle speeds is acceptable.(5)The analysis of wheel weight-time history curve obtained from different groups of strain data as input data shows that the proposed algorithm is sensitive to the number of sensors,sensor position and input strain data.Axle weight recognition result of measured strain data as input data indicate that installing the sensor only at the beam bottom of the midspan section can already provide sufficiently accurate results.There is no need to install sensors at each critical section,as this will limit the application of the BWIM system to wide bridges.(6)Based on the identification results of measured strain data of I-78 Bridge and I-459 Bridge,it is found that the results of the single axle weight identified by the vehicle with the wheelbase closer to the axle are not ideal,but the group axis recognition results have higher accuracy.Considering the actual situation,the weight of a single axle in a group axis that is close to each other can be approximated as an average distribution.Therefore,the axle load of a bridge vehicle identified by the algorithm proposed in this paper can meet the actual engineering application requirements.(7)The recognition results of the repeated running test of the calibrated vehicles in two bridges are very stable,which indicates that the algorithm proposed in this paper is a very effective and stable method for identifying the moving vehicle loads.