Research On Impact Load Inversion Based On Compressed Sensing And Model Analysis

Recently,the scholars on the impact force identification in aircraft primary structures have been undertaken in order to improve operation safety of these structures,reduce maintenance costs and prolong service life.How to obtain the impact force quickly and accurately is an important problem in the structural health monitoring.In this paper,the method of modal analysis based on compressive sampling data is proposed to solve the problem that traditional load inversion depends on structural mechanics model,which replaces the modal analysis method based on finite element analysis.A new method for the identification of impact load is presented based on the theory of compressive sampling,which includes the sparse model,the measurement matrix and the reconstruction optimization model.This method can overcome the problem that the amount of data of the stored and transmitted response signals is too large.Finally,a simulation example of a large aluminum plate is used to illustrate the impact force inversion method.The results show that the method can quickly and accurately identify the impact force on the large structure.The main contents of this paper include:The impact inversion based on Duhamel integral and modal superposition method is studied.The effects of sensor arrangement,load history,time step,modal order,Chebyshev order and regularization parameter on impact inversion were studied respectively.Firstly,the finite element model of aerospace aluminum plate is simulated by PATRAN/NASTRAN software.The structural parameters and response data are obtained by modal analysis and transient analysis.And then use the quadratic programming method to solve the optimization model on MATLAB.The impact force inversion method based on compressive sampling is studied.Firstly,the sparse representation of the signal under different sparse dictionaries like Chebyshev basis,Fourier basis,Discrete cosine basis is studied.And then we study the correlation between the different measurement matrices and the sparse base to obtain sufficient information to ensure the accuracy of the reconstruction,the influence of the selection of random measurement matrix and different compression ratios on the reconstruction accuracy of a single sensor response signal are discussed.The modal analysis method based on compressive sampling signals data is studied.Firstly,the feasibility of the modal analysis method based on the compressive sampling signal and the singular value decomposition is studied.The relationship between modal parameters and sampling parameters of the uniform sampled signal and the compressed sampled signal are studied,such as the length of the signal,the sampling period,the number of sampling points.So as to obtain the sampling parameters of the structure studied and to ensure the accuracy of modal analysis.At last,the feasibility of the impact load inversion method based on the compressive sampling signal and the modal analysis based on singular value decomposition is studied.Then,the algorithm of Quadratic Programming method and Basic Pursuit Denoising method based on compressive sampling signal is studied,and the compressed sampling signal used to directly obtain the structural impact load.At last,we discuss the influence of Gaussian matrix,0-1 uniformly distributed random matrix,Bemouli matrix and different compression ratios on impact load inversion.The simulation results show that the proposed method based on compressive sampling signal and modal analysis is effective.