Study Of Pseudo-Excitation Method For Random Vibration

Random vibration effect in the process of ground transportation, exhaust noise effect of the propulsion process, and aerodynamic noise effect on the aircraft in the process of high-speed flight receive increasing attention. Random response prediction and random loading identification is still the difficulty of vehicle dynamic environment study. A new calculation method, pseudo-excitation method (PEM) switches the calculation of stationary random vibration response to the calculation of stationary harmonic response, and uses the basic method of structure dynamics to treat the stationary random vibration response problem. It improves the efficiency, and also keeps the accuracy. The inverse pseudo-excitation method (IPEM) inheriting the high efficiency and high precision of PEM, can provide a clue to loading identification. In this paper, PEM is employed to study the problem of stationary random vibration response prediction and random loading identification, and the latter is the main content.An analysis model is built by finite element method, and a numerical simulation is presented to hold an accuracy and efficiency assessment among PEM, complete quadratic combination (CQC) and square root of the sum of squares (SRSS) methods in the calculation of random response power spectral density (PSD) of linear system. It proves the calculation amount of PEM is not sensitive to the number of modal superposition, and it is an accurate and high efficiency method. A simple experiment is designed to certify the feasibility and correctness of random response prediction calculation by PEM, also to analyze its anti-noise ability;The inverse pseudo excitation method (IPEM) for loading identification is mainly studied here. In different models of 3 DOF,9 DOF, and an finite elements model,with plenty of computer simulations, the influence by number and place of response measurement point ,damping ratio, precision of measurement of response, error of systematic parameter, and the noise of environments to loading identification is analyzed in detail, and some improved ways is given to reduce the error; several common algorithms are given to treat the ill-condition matrix, and the condition number weighted average algorithm is presented to handle the ill-condition transfer function matrix, and it is verified both by the computer simulations and experiments.Coherent and incoherent, two different types of random excitation experiments are designed to certify the feasibility and correctness of IPEM, the condition number weighted average algorithm is used to reduce the error caused by the ill-condition transfer function matrix, and the response measurement points is chose by the mean condition number of transfer function matrix, and it achieved good results; the error of the experiment is analyzed.