Modal Parameter Identification For Linear System Based On Improved Time-Frequency Slice Decomposition

System parameter identification can be divided into two research directions: time-invariant system parameter identification and time-varying system parameter identification.Among them,time-invariant system parameter identification has become mature,and a large number of commercial software has been successfully developed.In contrast,parameter identification of time-varying systems is a frontier research direction,but also the difficulty to be tackled urgently.In practical engineering,the mechanical properties of complex structures often change dramatically with time,so the research on parameter identification of time-varying systems has far-reaching engineering significance.Slice wavelet bases can analyze signals in any time-frequency range.Based on this,this paper introduces the method of slice wavelet bases into the field of parameter identification.Aiming at the problem of signal overlap which may occur in the case of time-varying structure,the original method of slice wavelet bases is improved by ridge extraction technology and adaptive slice domain recognition algorithm,and a new method based on slice wavelet bases is proposed.The system parameter identification method based on improved time-frequency slice decomposition firstly divides the slice interval adaptively in time-frequency domain according to the time-frequency energy distribution map of the response signal,and then reconstructs and restores the time-domain signal in each slice interval using inverse transform algorithm,thus realizing the decomposition of the response signal.The advantage of this method is that it can accurately separate the adaptive signals according to the time-frequency characteristics of the original signal,so as to improve the accuracy of subsequent parameter identification.Combining with the vibration equation,this paper deduces the parameter identification method based on improved time-frequency slice decomposition under two different excitations,impulse excitation and environmental excitation,and designs a simulation example to show its good identification results and anti-noise ability.