A Research On The Load Identification Based On Tikhonov Regularization And Its Application

Dynamic load identification is an important problem need to be solved urgently in modern engineering design. It is the basis of engineering structure reliability and safety design to determine exactly the dynamic loads acting. Also it has great significance for the structure fault diagnosis, life estimation and parameter identification and so on. However, because of the economic conditions or technical limitations, it is generally difficult to measuring the dynamic load directly in the practical engineering, but the response measurement of the system structure is relatively easy and accurate. Therefore, by using the structural dynamic response and the information of system to identify the dynamic load is an important method of indirect access. In this paper, on the basis of the traditional Tikhonov regularization methods, in view of its shortcomings, an iterative optimization load identification method is put forward. Then finite element and Model Reduction technique is applied to solve the dynamic response of the system in the process of iteration, to improve the efficiency of load iteration calculation on the promise of ensuring the accuracy of the solution. Finally, by using the iterative optimization load identification put forward in this paper, combined with a kind of the reduced and rigid-flexible coupled vehicle model, an inverse method to calculate the road excitation of vehicle is studied. The major research works in this paper are as follows:(1) An iterative optimization load identification method based on Tikhonov regularization technique is proposed. In the case of high noise level, the traditional Tikhonov regularization methods often show the excessive smooth side effects in the prcess of the load identification, affect the accuracy of the identification. In view of the problem, the iteration procedure based on Tikhonov regularization is used to modify the load signals in order to make the system real responses approach the expected signals. Theoretical analysis and numerical examples show that the presented method can retain a good anti-noise characteristic and improve the identification precision of the peak load place, the overall accuracy is improved.(2) Combined with finite element and model reduction techniques, the iterative optimization load identification method based on Tikhonov regularization technique studied in this paper is improved. After the comparison and analysis of reduced accuracy on the different model reduction method, appropriate reduction method is selected to reduce the extracted matrix system, and then a load identification of virtual iteration based on Tikhonov Regularization and model reduction is proposed. The new method can keep the accuracy of the identification, improve its application and the calculation efficiency of the iterative process.(3) Combined with an improved vehicle model, the application of the iterative load identification technique is studied to calculate the road displacement excitation of vehicle. By replacing a rigid sprung weight of the tradition multi-rigid-body vehicle models of 7 dofs with a flexible body of FEM, a rigid-flexible coupled model is constructed. The flexible body is attached to the rigid unsprung weight by means of linear spring. With the method studied in this paper, the load identification of virtual iteration based on Tikhonov Regularization and model reduction, the road displacement excitation can be calculated by the coupled vehicle system model and its response. It proves that the method is applicable to the road excitation identification of complex vehicle model, and has strong engineering application value.