Research Of Modal Parameter Identification Based On Rational Fitting Least Squares Complex Frequency Domain Method And Software Implementation

Vibration signal analysis is widely used in machinery, construction and other industries. In order to improve the integral result of long time vibration signal and the stabilization diagram of modal parameter identification, this paper studied on time domain integral analysis of vibration signal and modal parameter identification based on least squares complex frequency domain method.Firstly, the methods of removing the time domain signal trend items were compared. The method of dividing signal was proposed to remove the time domain signal integral trend items and to calculate integral result of long time signal. A simulation case was employed to validate this method. The results show that removing the time domain signal integral trend items in section is better than removing in whole signal.Secondly, the method of modal parameter identification based on the least squares complex frequency domain was researched. The different mapping range of frequency and the relationship of matrix polynomial order were indicated respectively using real coefficient matrix and complex coefficient matrix. The characteristics of identification and different working condition of two different methods respectively based on the positive power spectral of response and the transmissibility of response were discussed. The least squares complex frequency domain method was applied to both operational modal analysis. A simulation case was employed to validate these methods.Thirdly, owing to the fact that modal parameter could not be calculated from frequency response functions which were disturbed by noise after simply post-processing in time domain and frequency domain, the iteration method of rational function fitting was introduced to the least squares complex frequency domain method. A simulation case was employed to validate this method. The result shows that stabilization diagram is better in this method and is conducive to picking system poles.Finally, a vibration signal analysis system with DS-NET acquisition front-end was developed in Matlab platform. The software was suitable for time domain analysis, frequency domain analysis, and modal parameter identification. Both a case of modal analysis of an aluminum plate and a case of vibration analysis of slag pot carrier cab were carried by this software system.