| Accurate monitoring of structural dynamic loads is of great significance to ensure the safe operation of structures.In engineering practice,structural dynamic loads such as impact loads,wind loads and vehicle loads are often difficult to pick up directly.For this reason,indirect identification of dynamic loads has long been an important part of structural health monitoring.The dynamic load identification method usually uses the response information of the structure to invert the dynamic load of the structure,which belongs to the second type of inverse problem of structural dynamics.In the process of solving the load identification formula,the regularization method is introduced as a common means to improve the pathology of the identification equation.However,in the regularization method of existing load identification,the regularization used mostly belongs to vector regularization,which is suitable for dynamic load identification problems with small time span.In the vector regularization model,with the increase of the time span,the scale of load identification problems increases,and vector regularization is prone to dimensional disasters,which will bring obvious calculation efficiency degradation and is difficult to apply to dynamic load identification conditions for a long time.This paper focuses on the problem of dynamic load identification with a large time span,as follows:(1)The research status of dynamic force identification methods is reviewed.The methods of frequency domain method,time domain method,time-frequency domain method,timefrequency domain method,the suitability of the recognition equation,and the robustness of noise are compared,and the shortcomings of the existing methods are pointed out.On this basis,a dynamic load inversion method for linear structures under the regularization constraint of Frobenius norm is proposed.This method mainly adopts the response information under the time domain of the moving time window and organizes it into the form of a matrix sequentially,and introduces the Frobenius norm to solve the regular solution of the force identification equation.The proposed method can be regarded as an improvement over the traditional Tikhonov regularized force identification method,which not only retains the advantages of high recognition accuracy of the traditional Tikhonov regularized force identification method,but also greatly improves its identification efficiency.(2)The numerical case of cantilever beam was used to verify the effectiveness and feasibility of the proposed method.The discrete trigonometric function is selected as the load dictionary basis function,and the cantilever beam model under single impact load excitation and continuous impact load excitation is used for analysis,and the influence of noise level,measurement point combination,sampling frequency and other parameters on the proposed method is discussed.The results show that the proposed method has strong robustness and shows strong recognition accuracy under various working conditions.(3)Based on the numerical case of continuous beams to verify the efficiency of the proposed method in recognition efficiency,the force identification efficiency under different moving time window lengths is analyzed,and the influence of the change of moving time window length on the recognition accuracy is compared.Then,the proposed method is compared with the traditional Tikhonov regularized force identification method,and the superiority of the proposed method and the traditional Tikhonov regularized force identification method in recognition efficiency is verified.(4)Method verification through test case expansion.During the test,a cantilever beam model was built for testing,and the impact load excitation test under the cantilever beam was launched.Before the test,the finite element model was modified by modal test to refine the finite element model.Secondly,a single impact load test and a continuous impact load test are used for test verification.The test results show that the proposed method is effective,and has high recognition accuracy and recognition efficiency under the indicated working conditions. |
