Bridge Moving Load Identification Based On L₁ Regularization Method

With the increasingly serious problem of bridge structure safety in service,more and more attention has been paid to the research on health monitoring and condition assessment of bridge structure.As one of the main contents of health monitoring and condition assessment of bridge structures,moving load identification is of great significance to the operation,maintenance,maintenance,reinforcement and design of Bridges.The traditional load identification method is to directly obtain the moving vehicle weight on the bridge by weighing technology,but it is difficult to popularize due to the influence of sensor technology and cost.In this paper,based on the above situation,the indirect identification of the moving load on the bridge is studied by combining the dynamic response testing technique and dynamic load time-domain method.The main research contents are as follows:(1)Establish the identification equation of moving load in the time domain,and use L1 regularization method to solve it,and conduct numerical simulation verification.In this paper,based on the mapping principle of bridge linear system,the least square identification equation of moving load is established on the basis of simplified vehicle-bridge model.Considering the unsuitability of the solution,L1 regularization method is used to solve the system equation.The results show that the sparse solution can be compressed by L1 regularization,and the sparse solution can accurately invert the moving load.(2)Through the comparison of numerical simulation calculation,the influence of the initial conditions of the bridge in the load identification study is considered.Since the bridge is not always in a static state before the moving load,the modal space representation method of the initial condition is introduced to construct a new load identification equation.On this basis,numerical simulation is used to compare the recognition results of moving load with and without the influence of initial conditions.The results show that the traditional method can identify the moving load without the initial condition,but the error of identification with the initial condition is large,that is,the influence of the initial condition cannot be ignored in the study of the moving load identification problem.(3)A regularization identification method of moving load L1 with initial conditions was proposed and verified by numerical simulation.Firstly,the initial conditions are expressed in modal space and the load identification equation considering the initial conditions is constructed.In the face of the problems existing in the new system matrix,the new system matrix is processed based on the idea of average.A new method of load identification is proposed based on L1 regularization theory.The verification results show that the proposed method can better deal with the numerical differences between the column vectors of the system matrix,and it is better than the traditional sparse regularization method for the bridge moving load identification under unknown initial conditions.(4)Considering the complexity of the load identification problem,the identification capability of the proposed method was studied through numerical simulation and laboratory verification.Through numerical simulation,the load identification capability of the lifting method under different initial conditions and different axial load and speed of moving load is studied,and the influence of other factors such as different frequency and combination of measuring points is studied.The model of the single-span simply supported solid aluminum alloy beam bridge was established in the laboratory,and the static loading experiment,modal experiment and small ball experiment were carried out.The small ball experiment was the main experiment to verify the effectiveness of the proposed method.The experimental results show that the proposed method can accurately identify the moving ball on the laboratory bridge and has good practical adaptability.