Non-stationary Wind Characteristics And Structural Modal Parameter Identification Based On Field Measurements

With the rapid economic development and urbanization in recent years,the span and height of the structure are constantly breaking through,and the structure type is becoming complicated.For such structures,in-depth understanding of their load characteristics and response mechanisms has become a research hotspot in the engineering community.Relative to theoretical calculations and scale model experiments.The wind load and structural response obtained from the field measurement as first-hand information,which can be used to correct numerical simulation and scale model experiments,and it is a great significance for understanding correctly the wind load and structural response mechanism.Based on the health monitoring system of a bridge over sea and a large-span high-speed railway station,the wind speed data of the super strong typhoon "Mangkhut" and the structural response of the high-speed railway station building under the excitation of multiple load cases were obtained.Firstly,given the measured wind speed time historydata of typhoon at multiple measuring spots,the mean wind speed,wind direction,turbulent integral scale and fluctuating wind power spectrum were calculated,and the two important characteristics of fluctuating wind characteristic including the turbulent integral scale and fluctuating wind power spectrum,were investigated;The non-stationary characteristics of typhoon "Mangkhut" were investigated in detail,the time-varying mean wind speed of typhoon was identified,and the energy distribution characteristics of non-stationary wind in time and frequency domains were investigated by HHT and EPSD method.Secondly,the common noise reduction methods of time historydata signals were reviewed.A new noise reduction idea was proposed for non-stationary wind timehistory data,and the noise reduction methods of EEMD or CEEMDAN decomposition combined with permutation entropy had significant noise reduction influence by numerical experiments.Measured acceleration data of the steel structure station during train arrival,departure and pass,crowd-induced excitations and construction load excitation were analyzed.Investigation of the vibration response characteristics of the steel structure station under different test conditions were carried out.The study found that the vertical acceleration response frequency of the structure was multiplied by the train loading frequency.The response of the station under the excitation of train passing was significantly larger than that of the train entering the station,and they did not cause significant vibration;Equivalent Sinusoidal Peak Acceleration(ESPA)could better reflect the trend of acceleration amplitude,It is recommended to use this curve as a method to evaluate the comfort of the station.Based on the measured acceleration data of the high-speed railway station,the dynamic parameter identification of the large-span dense frequency structure was conducted,and it was found that the AMD method could not be accurately decomposed under the low-sampling-rate condition.To this end,an improved method was proposed,and combined with the VMD method,a variety of low-frequency modes were obtained.The E-RDT method was introduced to obtain the natural frequency and damping ratio of each order,and the relationship between amplitude and dynamic parameters was investigated.Finally,the major conclusions and the research points for future work were given.