Method for monitoring damage to high-rise building structures based on seismic background noise

Abstract

The present application relates to a kind of based on seismic background noise effectively monitor high-rise building structure damage monitoring method, belong to building structure monitoring technical field;With steps are (1) monitoring: using three-component seismograph long time continuity monitoring high-rise building seismic wave observation data, record vertical direction, north-south direction and east-west direction's ground micro motion time series;(2) data analysis: using sliding time window method to interference data is pretreated, obtain the original data of operable;With HVSR method determines the vibration, resonance frequency of building and observes selection amplification effect;Again, the mutual correlation analysis of HVSR spectrum obtained by calculation is carried out to systemically analyze the micro motion difference correlation between building floors, and to predict the structural damage of building. The method can monitor building vibration, floor micro motion difference, form a kind of new high-rise building structure maintenance monitoring scheme, scientific and reasonable, monitoring cost is low, high precision.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of based on seismic background noise effective monitoring high-rise building structure damage monitoring method, belong to building structure monitoring technical field. BACKGROUND

[0002] Seismic background noise refers to the persistent seismic signal wave recorded on broadband, generally consists of two parts, one is caused by environmental factors (such as wind, ocean movement, etc.) generally less than 1Hz seismic wave, the other is caused by human activities greater than 1Hz seismic wave. It is modified by the transmission medium in the process of propagation, therefore, by analyzing background noise can be indirectly analyzed the change of building structure through which the propagation. Seismic background noise analysis was first proposed by Kanai and Tanaka using background noise data analysis to calculate the site effect evaluation of major earthquake events. After Nakamura introduced the horizontal-vertical spectral ratio (HVSR) method in 1989, seismic background noise has gradually been valued by people, and its use has begun to constantly enrich and develop.

[0003] At present, many domestic and foreign researchers have conducted in-depth research by using seismic background noise as a medium and tool, but most of the research focuses on the characteristics of the modified medium of background noise, exploring the geological characteristics, velocity structure, etc. of a certain area. For example, Guo Zhi, Gao Xing team used seismic background noise imaging technology to obtain the crustal shear wave velocity structure of Tianshan and its surrounding areas, and then obtained the main driving force of Tianshan orogenic belt activation; Liu Chengguang, Hua Qingfeng team set up 100 mobile stations in Guangzhou to observe seismic background noise and obtain the distribution of sedimentary layers in the area, and then obtain the value representing the degree of damage to the site, providing reference data for urban disaster prevention and mitigation.

[0004] We found that there is less research on the analysis of super high-rise building structure by means of seismic background noise response, i.e. building structure monitoring. Although Wang Mingmin, Wang Yuan team tried to measure the horizontal and vertical seismic response characteristics of the building by using seismic waves, and roughly modeled to analyze the relationship between the natural vibration period of the building and the site characteristic period, and detected the safety of the building structure, but this is a rough modeling, and the reliability is low.

[0005] At present, the main method for measuring building structure is digital model method, shaking table method or real earthquake observation. Compared with these methods, building monitoring using seismic background noise can be applied to non-seismic zone, does not need digital or physical modeling, can reduce monitoring cost, and uses original building measurement, and the result is more practical. Using seismic background noise analysis needs enough vibration wave source, and ground pulsation generated by urban underground traffic can provide enough intensity wave source in each frequency band, so as to facilitate the promotion of building full-cycle structure monitoring in urban area, and provide monitoring technical support for building maintenance in the future. SUMMARY

[0006] The application aims to provide a monitoring method for effectively monitoring high-rise building structure damage based on seismic background noise, which adopts three-component seismograph to perform long series and continuous observation, and systematically analyzes the micro-motion difference correlation between building floors, so as to analyze the easily damaged floor structure of the whole building under the action of load.

[0007] In order to achieve the above purpose, the technical scheme adopted by the application is:

[0008] A monitoring method for effectively monitoring high-rise building structure damage based on seismic background noise, comprising the following steps:

[0009] (1) Monitoring: using three-component seismograph to continuously monitor high-rise building seismic wave observation data for a long time, recording the vertical direction, north-south direction and east-west direction micro-motion time sequence;

[0010] (2) Data analysis: using sliding time window method to pretreat interference data, obtaining operable original data; using HVSR method to measure the natural vibration and resonant frequency of the building and observe the amplification effect; then performing cross-correlation analysis on the calculated HVSR spectrum to systematically analyze the micro-motion difference correlation between building floors, and predicting the structure damage of the building.

[0011] The main data and data interference processing of step (2) refers to the vertical direction, north-south direction and east-west direction of the three-component seismograph record time series. These initial time series cannot be directly analyzed due to the influence of factors such as instrument response and occasional energy burst. The data preprocessing method of the HVSR method is as follows: ① Divide the three-component time series of the ground microseism into 1-hour long time segments, and there is no overlap between the segments; ② Remove the linear trend of the average value of the time segment and the instrument response in each segment; ③ Use the anti-trigger algorithm based on the specified range, STA time window is 1s and LTA time window is 30s, and the average amplitude ratio trigger threshold is 0.2<STA / LTA<2.5, to filter each time segment to prevent occasional energy burst and remove the marked transient interference signal.

[0012] The microseismic HVSR method is to preprocess the data stored by each station, obtain the microseismic data recorded by the station in multiple time windows, and perform fast Fourier transform on the microseismic data in each time window to obtain the frequency spectrum of the three components [V(t, f), N(t, f), E(t, f)]; Take the bandwidth coefficient as 40 and use the Konno-Ohmachi method to smooth the obtained frequency spectrum to facilitate the identification of important parameters such as dominant frequency; Calculate the HVSR value using formula (1), and combine the HVSR values calculated by each station into a curve and combine it into an HVSR result map for analysis, so as to quickly estimate the resonance frequency of the acting object and intuitively show the amplification effect.

[0013]

[0014] The peak value corresponding to the frequency generated simultaneously by each floor in the frequency spectrum is the natural frequency of the floor, and the peak value of each floor at the natural frequency reflects the selective amplification effect of the floor, while the peak value generated by individual floors at non-natural frequencies indicates the harmonic frequency of these floors.

[0015] The cross-correlation analysis of step (2) is a commonly used signal analysis method, which is used to process two HVSR signals corresponding to a station pair composed of two random stations. The corrcoef function of the Matlab program is used for calculation, and the specific calculation process is as follows:

[0016] For two random variables A, B, if each variable has N scalar observation values, the Pearson correlation coefficient is defined as follows:

[0017]

[0018] Where cov(A, B) is the covariance of random variables A and B, σA is the standard deviation of random variable A, and σB is the standard deviation of random variable B. The calculation of cov(A, B) can be referenced in the following formula:

[0019]

[0020] In the formula μ A Let A be the mean of the random variable A. Its calculation can be referenced in the following formula:

[0021]

[0022] Standard deviation σ A The calculation can be referenced from the following formula:

[0023]

[0024] When calculating the correlation coefficient, its absolute value is taken. There is currently no consensus in the statistical community regarding the meaning of the correlation coefficient, but it is generally considered that: an absolute value of 0.00-0.30 indicates a slight correlation between the two random variables; 0.30-0.50 indicates a real correlation; 0.50-0.80 indicates a significant correlation; and 0.80-1.00 indicates a high correlation.

[0025] Cross-correlation analysis was performed on the obtained HVSR signals, and the resulting correlation coefficient represents the linear correlation between the station and the two HVSR signals. Furthermore, since the HVSR signal is the result of integrated processing of the station's micro-motion signals, the obtained HVSR signal correlation coefficient can actually be considered as the correlation coefficient of the station's micro-motion signals.

[0026] The frequency used as the independent variable represents the frequency of the wave currently experienced by the analyzed floor. When the frequencies of the independent variables for a pair of stations are the same or nearly the same, it indicates that the floors corresponding to the two stations are basically under the influence of waves emitted by the same wave source. The cross-correlation coefficient within this range represents the linear correlation of the micro-motion signals exhibited by the two floors under the same wave influence. When the cross-correlation coefficient within this range is high, it indicates that the micro-motion generated by the two floors under the same influence has basically no large phase difference, that is, no large micro-motion difference is generated, the micro-motion difference between floors is small, and the corresponding floor shear stress is also relatively small.

[0027] By combining the calculated cross-correlation parameter diagrams to form a cross-correlation analysis diagram, the correlation of deformation micro-motions between floors and the shear force situation can be intuitively analyzed.

[0028] The meteorological data described in step (2) is mainly generated by the ground microseismic wave in the 0.3Hz-0.6Hz frequency band under the action of natural activities, and by investigating the meteorological data at the time of measurement, the change amplitude of the polarity parameter is related to some related parameters of the natural weather, and then the change amplitude of the polarity parameter is related to the load, displacement and other parameters to finally prove the relationship between the polarity parameter and the interlayer displacement angle and other important structural parameters.

[0029] The beneficial effects of the present application are:

[0030] The natural frequency of the building structure is quickly obtained by the HVSR spectrum obtained by calculation, and the phenomena of the selected amplification effect and harmonic resonance of the building can be directly observed; the microseismic differential correlation of the floor group is directly found by the cross-correlation image of each floor group, so as to analyze the easily damaged floor structure of the building under the action of load, which is scientific and reasonable, low in monitoring cost, high in precision, solves the problems existing in the existing monitoring technology, and is worth promoting. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 It is a test station point distribution diagram, wherein (a) is a plane point distribution diagram, and (b) is an elevation point distribution diagram;

[0032] Figure 2 It is an HVSR analysis diagram;

[0033] Figure 3 It is a main floor cross-correlation analysis diagram; DETAILED DESCRIPTION

[0034] The present application will be further described in detail by the following examples, which are only used to illustrate the present application and do not limit the scope of the present application.

[0035] 1 Sampling address, observation site and time

[0036] The Zhonghe Building of Tongji University in Shanghai is located near Shanghai Metro Line 10, and the ground microseismic generated by the subway can provide a wave source with sufficient intensity in each frequency band. This experiment set 22 observation points from -1 floor to 21 floor of the Zhonghe Building Figure 1 , and set 2 control observation points on the ground. The specific measurement station point distribution position is shown in Figure 1 , and the 21-day three-component seismic data collected from November 20, 2021 to December 11, 2021 is taken as an example to analyze the correlation between the horizontal-vertical frequency ratio spectrum, the polarization parameter, the cross-correlation function and the time and floor changes, and to calculate the predominant frequency and the correlation coefficient, and to comprehensively analyze the seismic response characteristics of the high-rise building, and to analyze the damage of the building structure.

[0037] 2 Data processing results

[0038] 2.1 Results of the Horizontal-Vertical Spectral Ratio (HVSR) Method

[0039] By slicing the generated HVSR image within important frequency bands, the HVSR analysis can be obtained as follows: Figure 3 ,Depend on Figure 3 It can be seen that the HVSR curves of all 24 stations exhibit peaks at approximately 0.4 Hz and 40 Hz. Furthermore, the peak value at 0.4 Hz increases with increasing floor level. Additionally, the difference between the HVSR values ​​at 0.4 Hz for floors one floor higher and lower increases with the overall number of floors. At 40 Hz, the peak-to-peak value changes are more irregular, and the difference in HVSR values ​​between floors no longer shows the clear positive correlation with floor level as it does at 0.4 Hz.

[0040] Figure 2 At frequencies of 1.6Hz, 2.4Hz, and 3.2Hz (multiples of 0.4Hz, 4, 6, and 8), except for the two stations on the first floor and the outdoor station, multiple small peaks appeared at other stations. The floors with the largest peaks were the 9th, 13th, and 5th floors, which were located in the middle of the building structure. The further away from the middle, and the closer to the top floor and the first floor, the smaller the peaks of the HVSR curves of the floors at these frequencies.

[0041] 2.2 Results of cross-correlation analysis

[0042] 2.2.1 Results of cross-correlation analysis of the main floors

[0043] The main structure of the building is divided into three floor groups: floors 2-7, 8-13, and 14-21 (low, middle, and high). The cross-correlation parameter diagrams of the station pairs (top and bottom floors of each floor group) and the station pairs (top floor 21 and bottom floor 2 of the main structure) are combined to form a cross-correlation analysis of the main floor levels. Figure 3 ;

[0044] Depend on Figure 3 It can be seen that the cross-correlation images of all stations in the main structure show a straight line image with a high cross-correlation coefficient, resembling a direct proportional function.

[0045] The absolute values ​​of the cross-correlation coefficients between the top and bottom floors of the lower-floor group are relatively small within the square region formed by the 0.4Hz-2Hz frequency range, mostly around 0. Furthermore, the linear graph of the direct proportional function also breaks within this range, with the absolute value of the coefficient dropping from approximately 0.9-1 to around 0.5-0.6.

[0046] The top floor and the bottom floor of the middle floor group have a small absolute value of the cross-correlation coefficient in the approximate square area composed of 10Hz-20Hz, most of which is around 0, and the straight line image of the positive proportional function also breaks in this interval, and the absolute value of the coefficient in this interval is basically around 0.4-0.6.

[0047] The cross-correlation image composed of the top floor and the bottom floor of the high floor group is basically the same as the image composed of the middle floor group, and the main difference is that the absolute value of the coefficient is further reduced, basically around 0-0.4.

[0048] The cross-correlation image composed of the top floor and the bottom floor of the building as a whole has the image characteristics of the low floor group and the middle-high floor group, but the frequency intervals of the two squares are increased, respectively around 0.3Hz-2Hz and 5Hz-20Hz, and the approximate cross-correlation coefficient is close to the coefficient of the low floor group and the high floor group image.

[0049] 3 Result analysis

[0050] 3.1 Analysis of horizontal-vertical spectral ratio (HVSR) method

[0051] (1) Building natural frequency monitoring: All stations have wave peaks at 0.4Hz and 40Hz, and a large number of research experience shows that 0.4Hz and 40Hz are close to or equal to the predominant frequency of the building structure, and the essence of the wave peak is that the rapid increase of structure microseism is caused by the resonance of seismic waves and building structure, and the main component of these seismic waves is Rayleigh wave. Therefore, 0.4Hz and 40Hz can also be considered as the natural frequency of the building structure.

[0052] (2) Building selective amplification effect monitoring: At 0.4Hz, the floor height increases, and the peak value increases accordingly. It can be seen that the building has a selective amplification effect on the 0.4Hz ground microseismic wave. The wave transmitted from the floor below is strengthened after resonance and then transmitted upward. The higher it is, the stronger the wave is, the more intense the resonance is, and the peak value difference between floors increases. At 40Hz, there is no such phenomenon, which may be because 40Hz is a natural frequency of the structure, not the main natural frequency of the building, and its vibration mode contributes little, so the selective amplification effect is not obvious.

[0053] (3) Building resonance monitoring: At 4, 6, 8 and other multiples of 0.4 Hz, the wave peaks of the HVSR curves of most floor stations appear. These larger frequency ground microseismic waves indicate that the building structure will exhibit resonance phenomena under the action of ground microseismic waves at even multiples of the main natural frequency of 0.4 Hz. The wave peaks of the stations on the 1st floor and outdoors are smaller or do not appear. These ground microseismic waves cannot form resonance with the ground, but can produce resonance with the building floor. The resonance wave source frequency is in the range of 1 Hz-5 Hz. In the work of Gutenburg on Rayleigh wave frequency range and wave source properties, it can be seen that resonance waves in this range are mainly produced by weaker human activities (human movement) and stronger natural activities (weather) and the like. This indicates that these resonance waves are more concentrated in the lower floors of the Zhongxiandao building due to the activities of the teachers and students in the building, while the lower floors of the building are mainly used for teaching. The upper floors are mainly used for college management, scientific research, and decorative sky gardens. Compared to the lower floors of the Zhongxiandao building, the activities of the teachers and students are more concentrated, and the resonance waves are also more concentrated in the middle and lower floors. Since the ground and the 1st floor cannot resonate with waves of this frequency, the lower floors are correspondingly constrained, resulting in weaker resonance intensity compared to the middle floors, which is manifested by the fact that the stations on the middle floors have the largest wave peaks at these resonance frequencies.

[0054] 3.2 Analysis of cross-correlation results

[0055] 3.2.1 Monitoring results of microtremor differences and shear forces between main floors

[0056] All cross-correlation images of station pairs appear a high coefficient straight line image of a positive proportional function, which indicates that in general, under the action of the same wave source, the microtremor correlation between the corresponding two floors of the station pair is good, and there will be no large microtremor difference, and there will be no large shear force between the main floors.

[0057] The positive proportional function image of the low floor group is interrupted at 0.4 Hz-2 Hz, and the cross-correlation coefficient in this range is poor, indicating that when subjected to wave sources at 0.4 Hz-2 Hz, the microtremor correlation between the floors of the low floor group is not ideal, and there may be a large microtremor difference between the floors. Especially at 0.4 Hz, which is near the natural frequency of the overall building, under the action of strong earthquakes at this frequency, the low floor group may produce a large microtremor difference between the floors, and further produce shear failure. However, considering that the wave sources in the frequency range above 1 Hz are mainly weak human activities such as human movement, and the intensity of such wave sources is generally not very large, it can be considered that the structure of the low floor group is safe, but it needs to be careful about the shear failure that may occur under the action of strong earthquakes at 0.4 Hz-1 Hz.

[0058] The high coefficient image of the middle floor group in the positive proportional function of 10Hz-20Hz is interrupted, the cross-correlation coefficient in this range is reduced, which shows that the micro-motion correlation of each floor in the middle floor group is reduced when affected by the wave source of 10Hz-20Hz, but the reduced cross-correlation coefficient still remains at 0.4-0.6, the micro-motion of the floor is between the real correlation and the significant correlation, and it can be considered that the structure of the middle floor group basically will not produce larger micro-motion difference, and the structure is safe.

[0059] The image of the high floor group is similar to that of the middle floor group, but the cross-correlation coefficient of the high floor group is reduced and remains at about 0-0.4, which can be considered that larger micro-motion difference will be produced between the floors when affected by the wave source of about 10Hz-20Hz. Considering that this frequency wave source is mainly generated by strong human activities, high-power scientific research equipment with vibration frequency at 10Hz-20Hz should be avoided in the floors of the high floor group, which will lead to long-term micro-motion difference between the floors and accelerate the fatigue of the building structure.

[0060] The image of the whole building has the characteristics of both the low floor group and the high floor group, and the cross-correlation coefficient is reduced at 0.4Hz-2Hz and 10Hz-20Hz, which shows that the micro-motion difference of the whole building changes roughly at 0.4Hz-2Hz, the lower floors of the low floor group produce micro-motion difference with the whole building, and at 10Hz-20Hz, the middle and high floors begin to produce micro-motion difference with the lower part of the building, and the higher the floor, the larger the micro-motion difference produced. And due to the difference in the reduced coefficients of the high floor group and the middle floor group, it can be seen that the higher the floor, the larger the micro-motion difference produced between the floors.

Claims

1. A monitoring method for effectively monitoring damage of a high-rise building structure based on seismic background noise, characterized by: It comprises the following steps: (1) Monitoring: using three-component seismometer to monitor the seismic wave observation data of high-rise building for a long time, recording the time series of ground micro-motion in vertical direction, north-south direction and east-west direction; (2) Data analysis: using sliding time window method to preprocess the interference data to obtain the original data; using HVSR method to measure the natural vibration and resonance frequency of the building and observe the amplification effect; then performing cross-correlation analysis on the calculated HVSR spectrum to systematically analyze the micro-motion difference correlation between the floors of the building and predict the structural damage of the building; The cross-correlation analysis in step (2) is used to process the two corresponding HVSR signals of a station pair composed of two random stations, and the corrcoef function of Matlab program is used for calculation, and the specific calculation process is as follows: For two random variables A and B, if each variable has N scalar observation values, the Pearson correlation coefficient is defined as follows: wherein, is the covariance of random variables A, B, is the standard deviation of random variable A, is the standard deviation of random variable B, The calculation of refers to equation (3): ， In the formula is the mean of the random variable A, which is calculated with reference to equation (4): ， Standard deviation The calculation reference formula (5): 。 2. The method according to claim 1, wherein the method is characterized by: The sliding time window method in step (2) for preprocessing interference data refers to the time series of ground micro-motion in vertical direction, north-south direction and east-west direction recorded by three-component seismometer. These initial time series cannot be directly analyzed due to the influence of instrument response and occasional energy burst factors. The data preprocessing method using HVSR method to process several interferences is as follows: ① divide the three-component time series of ground micro-motion into 1-hour time segments, and there is no overlap between segments; ② remove the linear trend of the average value of the time segment and the instrument response in each segment; ③ use the anti-trigger algorithm based on the specified range, STA time window is 1s and LTA time window is 30s, average amplitude ratio trigger threshold is 0.2<STA / LTA<2.5, to filter each time segment to prevent occasional energy burst and remove the marked transient interference signals.

3. The method according to claim 2, wherein the method is characterized by: The HVSR method is to preprocess the data stored by each station to obtain the micro-motion data recorded by the station in multiple time windows. Fast Fourier transform is performed on the micro-motion data in each time window to obtain the frequency spectrum in [V(t,f), N(t,f), E(t,f)] three components; take the bandwidth coefficient as 40 and use Konno-Ohmachi method to smooth the obtained frequency spectrum, calculate the HVSR value by formula (1), and combine the HVSR values calculated by each station into a curve to form an HVSR result graph for analysis, so as to quickly estimate the resonance frequency of the vibrating object and intuitively show the amplification effect 。 4. The method of claim 1, wherein the method is characterized by: The meteorological data used is mainly for the ground micro-motion wave in 0.3Hz-0.6Hz frequency band generated under the action of natural activity. By investigating the meteorological data at the time of measurement, the change amplitude of the polarity parameter is related to some related parameters of natural weather, and then the change amplitude of the polarity parameter is related to the load and displacement to finally prove the relationship between the polarity parameter and the interlayer displacement angle structure important parameter.

Images