Time-Varying Cable Force Identification Based On Generalized S-Transform Transmissibility

For bridge structures in the service period,time-varying cable force identification is the basis of cable limit state safety assessment and fatigue cumulative damage evaluation.The traditional cable force measurement based on the vibration method cannot recognize the varying real-time cable force.To solve this,a generalized S-transform transmissibility method is proposed and applied to the identification of time-varying cable forces.The method provides a new idea for time-varying cable force identification in practical engineering.The main contents of this paper are included as follows:The improved method of S transform and the influence of different parameter selection on window function are studied.Then a parameter optimization method based on clustering criterion is given for the generalized S-transform with two parameters.Compared with the traditional time-frequency analysis method,the generalized Stransform with optimized parameters for different signals retains the excellent characteristics of multi-resolution,lossless reversibility and higher time-frequency aggregation.With optimized parameters,the analysis results of the generalized Stransform are more consistent with the requirements of non-stationary signal analysis.A time-frequency analysis method of generalized S-transform-based transmissibility is proposed.Combined with the advantage of generalized S-transform in time-frequency analysis.the concept of transmissibility is extended to time-frequency domain.The method is based on auto-and cross generalized S transform spectra.Based on the H1 estimation model,the calculation method of the generalized S-transform-based transmissibility is given.It’s a time-frequency analysis method.The analysis results can be directly calculated from two sets of responses in different positions of the results,independent of the input hypothesis.Due to the randomness of noise signal in practical application,it is not related to other signals in the time-frequency domain,so the transmissibility calculation based on the generalized S transform shows a good anti-noise ability.In this paper,the vibration acceleration response of the same time-varying structure is obtained by constructing the simulation model of the three-degree-of-freedom time-varying structure excited by white noise.The analysis results of the response signals show that the time-varying characteristics of the structure can be better revealed by the method based on the generalized S-transform transmissibility,and the time spectrum has obvious continuous energy ridges.Compared with other time-frequency analysis methods,the generalized S-transform transfer transmissibility has better noise suppression characteristics for signals with the same SNR.In view of the problem that the traditional method of cable force identification based on frequency method can only identify the average cable force in the past period of time,but can not identify the cable force in real time,the generalized S-transform transmissibility method is applied to the identification of time-varying cable force.Under the simulation conditions,the generalized S transform transmissibility was calculated by two sets of vibration acceleration response signals at different positions of the cable,and then the ridge line was extracted from the generalized S transform transmissibility time spectrum by using the dynamic programming method to realize the identification of the first-order instantaneous frequency of the cable,thus the time-varying cable force was obtained.Under the experimental conditions,due to the impact excitation,the acceleration response signal gradually attenuates in the research time range,the signalto-noise ratio gradually decreases,and the influence of noise gradually becomes obvious,thus the recognition effect at the end of the signal will gradually deteriorate.Compared with the wavelet transform and the generalized S transform,the transmissibility of the generalized S transform can suppress the background noise better,so it has a more accurate effect of cable force identification.