The Research Of Modal Parameters Identification For Vibrant System

The estimation of vibrant system's modal parameters was an important part of modal analysis. Using vibrant system's measurements constructed the system's modal model and estimated the modal parameters from the model were the essences of modal parameters identification. The dynamic behavior of system can be described by the modal parameters, so estimated the modal parameters fast and exactly can give a important academic guidance to the analysis of system's dynamic behavior, the diagnoses of vibrant failure and the optimize of vibrant system's dynamic behavior.For now, the researches of modal parameters'estimation algorithms were mostly for frequency domain and time-frequency domain. This paper performed the researches of modal parameters'estimation after learning modal analysis theory and the existent estimation algorithms, mainly including following three aspects.(1) A new method of modal parameters identification improved on the least-square complex frequency-domain (LSCF) by the wavelet transform. The limitations that existed in the application of LSCF for the frequency response function (FRF) with high damping ratio and closely spaced modes were analyzed. A new method of choosing appropriate wavelet scale parameters for application the wavelet transform on the FRF was put forward. This method can decouple the modes which cannot be accurately operated by the LSCF and identified the modal parameters of them.(2) A self adaptive method based on dual-tree complex wavelet package transform. Using continuous wavelet estimated the modal parameter required the operator identify the wavelet ridgeline by man-made operation, so the precision of estimation was not accurate. The method based on dual-tree complex wavelet package transform can decouple the modes of vibrant system by self adaptive algorithm, dispensed with the man-made operation.(3)An optimization algorithm of the overcomplete matching pursuit atom dictionary. Using the matching pursuit method estimated the vibrant system's modal parameters had a huge amount of calculation. The optimization algorithm based on the character of modal parameters reduced the atomic number in overcomplete atom dictionary, thereby reduced the amount of calculation.