Multiple-input Multiple-output Frequency Domain Orthogonal Polynomials Modal Parameter Identification Method

Experimental modal analysis is the whole process to establish mathematic model of describing structure's vibration characteristic based on experiment. It comprises experiment and modal analysis. Along with the development of modern engineering technique, various structure dynamics problems are increasingly standing out. For instance, structure sensitive analysis, amending model of structure dynamics and system dynamic simulating control system dynamic require high accuracy and being widely used of experimental modal analysis. Multi-Input Multi-Output (MIMO) Frequency Response Function techniques appear in this case.The MIMO system modal parameter identification methods based on Frequency Domain Orthogonal Polynomial algorithm is stressed and discussed. Multiple input modal analysis has two problems to solve: the estimation of Frequency Response Function (FRF) and the global parameter identification of Multi-Input Multi-Output(MIMO) system. Three classical Frequency Response Function estimation methods H₁ H₂ and H_v are discussed firstly. By comparing the noise disposal manner andapplication areas and estimation errors of those methods, the advantages of H_vmethods are showed. Secondly, the MIMO system modal parameter identification methods based on Frequency Domain Orthogonal Polynomial algorithm is formed. Based on mathematics model of MIMO system, it adopt Orthogonal Polynomial to replace nature polynomial to fit FRF and identify system modal parameter. It provides idiographic methods of reducing influences of modals besides analyzed frequency belt and several criterions of model order determination. The forming of Orthogonal Polynomial matrix is reasoned. The MIMO system modal parameter identification methods based on Frequency Domain Orthogonal Polynomial algorithm is realized by MATLAB program. Finally, by simulation and experiment, the correctness of the program and the advantages of the MIMO system modal parameter identification methods based on Frequency Domain OrthogonalPolynomial algorithm were proved.The MIMO system modal parameter identification methods based on Frequency Domain Orthogonal Polynomial algorithm own some advantages. It reduces the ill conditions of the coefficient matrix in curve fitting and curtails orders of matrix operation greatly and fully uses all the information of experiment and gains all modal parameters. It improves the identification accuracy and the consistency of identified modal parameter and reduces man-made interference. It can also dispose of high resistance, dense models. The method can meet the requirement of high accuracy and being widely used of experimental modal analysis.