Method for surveying internal structure of structure, and system thereof

Abstract

To provide a method for surveying an internal structure of a structure using micro-vibrations, in particular, a structure to which human activity derived vibration is steadily applied, and to provide a system of said method.SOLUTION: There is provided a method for surveying an internal structure of a structure having an upper surface to which walking vibration is steadily applied. A structure is provided with sensors and micro-vibrations being temporal changes in vibration are measured, and a time window is extracted in which walking vibration from the micro-vibrations is applied, and then, auto-correlation analysis is conducted regarding vertical direction vibration that is included in the walking vibration inputted to an internal part of a structure, from the upper surface, and that is reflected, thereby surveying an internal structure of a structure. A system for surveying an internal structure of a structure comprises: a plurality of sensors for measuring micro-vibrations applied to a structure; and an arithmetic logical unit for surveying the internal structure of the structure from the micro-vibrations. The arithmetic logical unit extracts a time window in which walking vibration from the micro-vibrations is applied, and then, conducts auto-correlation analysis with regard to vertical direction vibration included in the walking vibration that is inputted to the internal part of the structure from an upper surface, and that is reflected.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a method and system for investigating the internal structure of a structure using noise vibrations, and more particularly to a method and system for investigating the internal structure of a structure that is constantly subjected to vibrations resulting from human activity. [Background technology]

[0002] In the exploration of ground and structures, a method is known in which elastic waves passing through the target area are measured using velocity sensors and acceleration sensors arranged to surround the target area, and the velocity distribution of the elastic waves within the target area is calculated and inverted to perform structural exploration. For example, in seismic reflection exploration, vibrations artificially generated by a vibroseis or the like are propagated from the surface to underground, and the waves reflected at the structural boundaries (reflected waves) are observed by sensors installed on the surface or in boreholes to image the underground structure. Since a known input signal is used, the signal-to-noise ratio is high, and underground structures can be imaged with high resolution.

[0003] Patent Document 1 also discloses a method for inspecting the presence or absence of damage or the like inside a concrete structure from the surface of the concrete using multiple elastic wave detection sensors. Specifically, elastic wave detection sensors are arranged at multiple points on the boundary surrounding the inspection area on the surface of the inspection part of the concrete structure to be inspected, and striking tools and striking force detection sensors that detect the striking force corresponding to each striking tool are arranged at multiple points where no elastic wave detection sensors are arranged. Then, the propagation time of the elastic wave from the oscillation point to the receiving point is calculated based on the rising detection point of the smoothed waveform of the striking force measurement waveform obtained by the striking force detection sensor and the rising or falling detection point of the smoothed waveform of the measurement waveform generated based on the measurement data obtained by each of the elastic wave detection sensors. This is subjected to a fast Fourier transform (FFT) to obtain the elastic wave speed from the dominant frequency, the first to fifth frequencies, and their amplitudes. Then, the method claims that a simple diagnosis of the condition inside the concrete structure can be performed by comparing the fundamental waveform generated based on the fundamental elastic wave data obtained when the concrete structure is in a sound state with the above-mentioned measured waveform. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-96197 Summary of the Invention [Problem to be solved by the invention]

[0005] Meanwhile, in fields such as earthquake observation, attempts are being made to monitor the time change of the crustal structure from the time change of the autocorrelation function (ACF) of natural noise tremors. It is considered to apply such a method to the exploration of structures, etc., but unlike the method using tomography analysis as disclosed in the above-mentioned cited document 1, it uses noise tremors as an input signal, so there is no need to install a vibration source and the system can be simplified, but since the input signal is unknown, the S / N ratio cannot be made high.

[0006] The present invention has been made in consideration of the above-mentioned circumstances, and an object of the present invention is to provide a method and system for exploring the internal structure of a structure using noise vibrations, in particular, the internal structure of a structure that is constantly subjected to vibrations resulting from human activity. [Means for solving the problem]

[0007] The method for exploring the internal structure of a structure according to the present invention is a method for exploring the internal structure of a structure having an upper surface to which walking vibrations are constantly applied, and is characterized in that a sensor is attached to the structure to measure micro-vibrations, which are the time changes in vibration, and a time window in which the walking vibrations are applied is extracted from the micro-vibrations, and the internal structure of the structure is explored by performing an autocorrelation analysis on the vertical vibrations contained in the walking vibrations that are input into the interior of the structure from the upper surface and reflected.

[0008] According to this feature, even if there is only minor vibration, by extracting a time window that includes walking vibration, it is possible to explore the internal structure of the ground or a structure.

[0009] The above-mentioned invention may be characterized in that it includes a step of integrating the vertical vibration to separate signals. With this feature, it is possible to increase the S / N ratio of the signal and to perform exploration of the internal structure of the ground or a structure.

[0010] In the above-mentioned invention, a frequency Fp having a maximum power spectrum and a frequency Fs having a second maximum power spectrum in a frequency spectrum of an envelope of a waveform obtained by filtering the measured noise vibration may be obtained, and the time window is extracted when the frequency Fs is a harmonic of the frequency Fp and the power spectrum of the frequency Fp is sufficiently larger than the power spectra of other frequencies. According to this feature, the time window in which walking vibration is applied can be easily extracted, and the internal structure of the ground or a structure can be explored.

[0011] In the above-mentioned invention, the extraction of the time window may be performed based on a signal from a detection means for detecting the entry and exit of a pedestrian on the upper surface. According to this feature, it is possible to efficiently extract a time window in which walking vibration is applied, and then to perform an internal structure exploration of the ground or a structure.

[0012] In the above-mentioned invention, the structure may be a ground with a paved surface. Also, the structure may be a bridge. According to such a feature, vertical vibration can be measured relatively clearly, and the internal structure of the ground or the structure can be explored.

[0013] In the above-mentioned invention, the sensor may be a velocity sensor or an acceleration sensor. According to this feature, it is possible to measure the vertical vibration with high accuracy and to perform an exploration of the internal structure of the ground or a structure.

[0014] In addition, the internal structure exploration system of the present invention is an internal structure exploration system for a structure having an upper surface to which walking vibrations are constantly applied, and is equipped with a sensor that measures microvibrations, which are the time changes in vibrations applied to the structure, and a calculation unit that explores the internal structure of the structure from the microvibrations, and is characterized in that the calculation unit extracts a time window during which the walking vibrations are applied from the microvibrations, and then performs an autocorrelation analysis on the vertical vibrations contained in the walking vibrations that are input into the interior of the structure from the upper surface and reflected.

[0015] According to this feature, even if there is only minor vibration, by extracting a time window that includes walking vibration, it is possible to explore the internal structure of the ground or a structure.

[0016] In the above-mentioned invention, the calculation unit may perform a process including integrating the vertical vibration and performing a process for signal separation. With this feature, the S / N ratio of the signal can be increased, and the internal structure of the ground or a structure can be explored.

[0017] In the above-mentioned invention, a frequency Fp having a maximum power spectrum and a frequency Fs having a second maximum power spectrum in a frequency spectrum of an envelope of a waveform obtained by filtering the measured noise vibration may be obtained, and the time window is extracted when the frequency Fs is a harmonic of the frequency Fp and the power spectrum of the frequency Fp is sufficiently larger than the power spectra of other frequencies. According to this feature, the time window in which walking vibration is applied can be easily extracted, and the internal structure of the ground or a structure can be explored.

[0018] In the above-mentioned invention, the apparatus may further include a detection means for detecting the entry and exit of a pedestrian onto the upper surface, and the calculation unit may extract the time window based on a signal from the detection means. According to this feature, it is possible to efficiently extract the time window in which walking vibration is applied, and then to perform an internal structure exploration of the ground or a structure.

[0019] In the above-mentioned invention, the structure may be a ground with a paved surface. Also, the structure may be a bridge. According to such a feature, vertical vibration can be measured relatively clearly, and the internal structure of the ground or the structure can be explored.

[0020] In the above-mentioned invention, the sensor may be a velocity sensor or an acceleration sensor. According to this feature, it is possible to measure the vertical vibration with high accuracy and to perform an exploration of the internal structure of the ground or a structure. [Brief description of the drawings]

[0021] [Figure 1] FIG. 1 is a block diagram of an internal structure exploration system according to one embodiment of the present invention. [Diagram 2] FIG. 2 is a flow chart showing an internal structure exploration method according to one embodiment of the present invention. [Diagram 3] FIG. 11 is a flow chart showing a method for determining whether or not walking vibrations are applied. [Figure 4]FIG. 13 is a diagram showing waveform processing for determining whether or not walking vibration is applied. [Diagram 5] This figure shows examples of waveforms measured by (a) miscellaneous vibration, (b) impact vibration, and (c) walking vibration. [Figure 6] FIG. 13 is a diagram showing stacked autocorrelation waveforms caused by noise, impact vibration, and walking vibration. [Figure 7] 13 is a diagram comparing the correlation coefficient of the autocorrelation waveform due to impact vibration with the autocorrelation waveforms due to noise and vibration and walking vibration. [Figure 8] 13 is a diagram showing the determination of segments to which walking vibration is applied for (a) miscellaneous vibration, (b) impact vibration, and (c) walking vibration. [Figure 9] 13 is a diagram in which some segments of noise and vibration are replaced with walking vibrations and segments to which walking vibrations are applied are determined. FIG. [Figure 10] FIG. 1 is a flow diagram showing an example of a method for exploring the internal structure of the ground. [Figure 11] FIG. 1 shows autocorrelation waveforms for each 20-second segment for (a) noise vibration, (b) impact vibration, and (c) walking vibration. [Figure 12] 1A is a diagram showing the estimated results of changes in propagation velocity for (a) noise and vibration, (b) impact vibration, and (c) walking vibration. [Figure 13] FIG. 13 is a diagram showing the standard deviation of the change in propagation velocity due to noise and walking vibration. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0022] An internal structure exploration method and system according to one embodiment of the present invention will be described below. First, the internal structure exploration system will be described.

[0023] As shown in Fig. 1, an internal structure exploration system 10 includes a sensor 3 that is installed in a structure 1 and can measure noise vibrations, which are time-varying vibrations in the structure 1, and a calculation unit 4 that performs analysis of the noise vibrations obtained from the sensor 3. Note that multiple sensors 3 may be provided. The mode of the sensor 3 can be optimized depending on the type of the structure 1, etc. Note that when multiple sensors 3 are provided, the following analysis can be performed for each sensor 3.

[0024] The structure 1 is a structure, mainly a building, and has an upper surface 2 to which walking vibrations are constantly applied. In other words, the upper surface 2 can be, for example, a road along which people walk or a passageway of a structure. As such a structure, it is preferable to use a structure having an upper surface 2 to which walking vibrations are clearly input, such as the ground of a paved road or a bridge.

[0025] On the other hand, the internal structure exploration system 10 measures and analyzes the micro-vibrations of the structure 1 as described above, but uses walking vibrations as a vibration source for analysis. Vibrations originating from human activities such as walking vibrations are easily attenuated, and are generally difficult to use for structural analysis using reflected waves of vibrations. Therefore, in the internal structure exploration system 10, vertical vibrations are measured and used for analysis. If the vertical vibrations are input into the structure 1 and reflected, the shortest path from the vibration source to the sensor 3 is the one with the least attenuation. Also, as described later, if the vibrations are repeatedly added, they can be integrated, and the S / N ratio can be further increased. In other words, even if the vibrations originating from human activities, which are easily attenuated, can be measured as relatively clear vibrations. The sensor 3 is capable of measuring vertical vibrations including such reflected vibrations as micro-vibrations. For example, it is preferable to use a speed sensor or an acceleration sensor, since it is possible to measure the vertical vibrations with high accuracy.

[0026] The sensor 3 is connected to a calculation unit 4. The calculation unit 4 analyzes the noise vibrations obtained from the sensor 3 and explores the internal structure of the structure 1. As the calculation unit 4, for example, a personal computer can be suitably used, and it can be equipped with appropriate software for calculation.

[0027] Next, an internal structure exploration method using the calculation unit 4 will be described with reference to FIGS.

[0028] 2, the calculation unit 4 first measures the noise disturbance (S1). That is, the calculation unit 4 measures and records the noise disturbance in the vertical direction using the sensor 3. Next, the calculation unit 4 divides the measured noise disturbance into a plurality of time windows (segments) (S2), and extracts the time windows to which walking vibration is added (S3).

[0029] As shown in Fig. 3, in the extraction of time windows (S3), each of the waveforms of the noise vibrations divided into the time windows is filtered by an arbitrary band-pass filter (S11). It is preferable to adjust the pass band of the band-pass filter to a frequency band where walking vibrations are more clearly observed, taking into account the installation environment and the measuring equipment. For example, it can be set to 60 to 90 Hz to suppress vibrations caused by vehicle movement.

[0030] Next, the envelope of the filtered waveform is calculated (S12), and the frequency spectrum of the envelope is obtained by frequency decomposition using a Fourier transform (S13).Furthermore, the frequency Fp of the frequency spectrum with the maximum power spectrum (PS) and the frequency Fs of the frequency spectrum with the second maximum power spectrum (PS) are obtained (S14).

[0031] Then, if the power spectrum from which the frequency Fp was obtained is sufficiently larger than the power spectrum at other frequencies, and the frequency Fs is approximately twice the frequency (harmonic) of the frequency Fp (S15; Yes), it is determined that the waveform from which the envelope was obtained is dominated by walking vibration. That is, by estimating the walking period of a pedestrian from the period of the envelope of the observed waveform, it is possible to determine that walking vibration was imparted to the time window from which this waveform was obtained (S16). For example, the frequency of a wave in which a signal is observed three times per second is dominated by a frequency that is a multiple of 3 Hz. On the other hand, if at least one of the power spectrum or the frequency of the frequency Fp does not fall under these conditions (S15; No), it is determined that walking vibration was not imparted to this time window (S17). Then, this determination is repeated for a plurality of time windows, and a time window determined to have walking vibration imparted is extracted.

[0032] For example, as shown in FIG. 4, the measured noise vibrations (FIG. 4(a), including walking vibrations here) are filtered to obtain an envelope (FIG. 4(b)), and frequency decomposition is performed to obtain the frequency spectrum of the envelope (FIG. 4(c)). This allows Fp and Fs to be specified and judged as described above.

[0033] 2 again, the noise vibrations in the extracted time window are subjected to autocorrelation analysis such as signal separation (S4), and the internal structure of the structure 1 is explored, for example, by imaging the internal structure (S5). For example, the change in the propagation speed of the reflected wave obtained from the autocorrelation function by the autocorrelation analysis can be estimated to detect a change in the physical properties of the internal structure of the structure 1. Such autocorrelation analysis (S4) and exploration of the internal structure (S5) are well known, and therefore detailed description thereof will be omitted.

[0034] As described above, even with noise and vibration, by extracting a time window that includes walking vibration and analyzing the vertical vibration included in the walking vibration, the S / N ratio can be increased and the internal structure of the structure can be explored. In particular, by measuring the vertical vibration, it is possible to clearly obtain the reflected wave that uses the walking vibration caused by walking near the sensor 3 as the input.

[0035] It is also preferable to perform signal separation by integrating vertical vibrations for multiple time windows of noise vibrations extracted so as to include walking vibrations that will repeatedly cause vertical vibrations. By integrating multiple walking vibrations, a signal that has statistically certain characteristics as a reflected wave of walking vibration can be obtained, and a signal with a higher S / N ratio can be easily obtained.

[0036] Also, it is preferable that the extraction of the time window in which walking vibration is applied is performed based on a signal from a detection means that detects the entry and exit of pedestrians on the upper surface 2. For example, if it is within a time period in which it is estimated that pedestrians are walking on the upper surface 2 based on a signal from a pedestrian detection means such as an infrared detector, many time windows in which walking vibration is applied can be extracted, and if it is outside that time period, they may be excluded from the extraction. In other words, the extraction of the time window in which walking vibration is applied can be performed efficiently.

[0037] [Verification test] The validity of the above-described method for investigating the internal structure was verified as follows.

[0038] First, as shown in Figure 5, miscellaneous vibrations not including walking vibrations (Figure 5(a)), hammer impact vibrations (Figure 5(b)), and walking vibrations (Figure 5(c)) were measured. In detail, the sensor 3 of the internal structure exploration system 10 was installed on a paved ground surface and measurements were taken. Miscellaneous vibrations were measured for about 30 minutes with no specific input to the ground surface near the sensor 3. Impact vibrations were measured by hitting the ground surface several centimeters away from the sensor 3 with a hammer once every 20 seconds and measuring the reflected waves. Walking vibrations were measured by passing near the sensor 3 by walking or running once every 20 seconds and measuring the reflected waves. Each measurement was recorded at a sampling rate of 200 Hz.

[0039] For each sampled record, spline interpolation was performed in the time direction from 200 Hz to 1000 Hz to obtain each waveform. The obtained waveform was divided into segments every 20 seconds, and the autocorrelation function was calculated for each segment and stacked (averaged).

[0040] As shown in Figure 6, the autocorrelation waveforms of each waveform obtained were compared. Figure 7 also shows the correlation coefficients of the autocorrelation waveforms of walking vibration and miscellaneous vibrations not including walking vibration with respect to the autocorrelation waveform of hammer impact vibration. It can be seen that the autocorrelation waveform of walking vibration is closer to the autocorrelation waveform of hammer impact vibration than the miscellaneous vibrations not including walking vibration. Generally, in autocorrelation analysis using waves reflected from underground, the reflected waves of hammer impact vibration are used as ideal reflected waves. In other words, it can be said that the reflected waves from walking vibration are close to ideal reflected waves for use in autocorrelation analysis. Furthermore, it is suggested that clear reflected waves can be obtained by using walking vibration.

[0041] Figures 8(a), (b), and (c) show the waveforms measured for the noise vibrations not including the walking vibrations measured above, the hammer impact vibrations, and the walking vibrations. Furthermore, for each segment into which each waveform is divided, whether or not walking vibrations are included is judged in the same manner as above, and then a time window is extracted (S3). The extracted segments (determined to have been given walking vibrations) are shown in gray, and the others in black. As a result, for the waveform (c) regarding walking vibrations, all segments were determined to have been given walking vibrations, and for the other waveforms, no segments were judged to have been given walking vibrations.

[0042] In addition, as shown in Figure 9, a virtual waveform was created by replacing five segments of the waveform of noise vibration that does not include walking vibration with segments of the waveform of walking vibration. For this waveform, segments containing walking vibration were extracted using the same method as above. As a result, the only segments that were determined to contain walking vibration and were extracted were the five segments shown in gray where the waveform was replaced with walking vibration.

[0043] From these findings, it can be said that the time window to which walking vibration was applied was correctly determined and extracted by the above-mentioned method.

[0044] Next, we performed an exploration of the internal structure of the ground using the autocorrelation waveform obtained from the reflected waves measured in the extracted time window. Here, we detected changes in the physical properties of the structure by estimating changes in the propagation speed of the reflected waves.

[0045] In detail, as shown in Fig. 10, first, the reflected wave is divided into a number of segments, and the autocorrelation function of the reflected wave is calculated for each segment (S21). Next, the autocorrelation waveform is filtered using a bandpass filter (S22), and the waveform of each segment is obtained and used as the target waveform (S23). Furthermore, the average waveform of all segments is calculated and used as the reference waveform (S24). Then, each target waveform is stretched or shrunk, and the stretch / shrunk ratio that has the best correlation with the reference waveform is calculated (S25). The calculated stretch / shrunk ratio is multiplied by -1 to estimate the rate of change of speed (S26). This is a method called the stretching method, which estimates the degree to which each target waveform has changed relative to the reference waveform by stretching or shrinking the waveform.

[0046] Figures 11(a), (b), and (c) show the autocorrelation waveforms (60 to 90 Hz) for every 20-second segment for the noise vibrations not including the walking vibrations measured above, the hammer impact vibrations, and the walking vibrations.

[0047] Furthermore, Fig. 12(a), (b) and (c) respectively show the propagation velocity change rate of the autocorrelation waveform for every 20-second segment for the noise vibration not including walking vibration, hammer impact vibration and walking vibration measured above. The black solid line shows the estimated velocity change rate, and the background color shows the correlation coefficient of the target waveform and the reference waveform for the velocity change rate. The estimated velocity change rate is the velocity change rate that maximizes the correlation coefficient. Since the observation time is short, it is considered that there is no change in the structure, and ideally, it is expected that no velocity change will occur. And, in the exploration using the hammer impact vibration, the velocity change rate of the autocorrelation waveform was almost 0. On the other hand, in the exploration using the noise vibration not including walking vibration and the walking vibration, an apparent velocity change occurred in the autocorrelation waveform.

[0048] Figure 13 shows the standard deviation of the speed change rate of the autocorrelation waveform for both noise and vibration that does not include walking vibration. As mentioned above, ideally there should be no speed change. Therefore, it can be said that the autocorrelation waveform due to walking vibration, which has a small standard deviation, provides a more stable reflected wave than the noise and vibration, which has a large standard deviation.

[0049] As described above, according to this embodiment, even if the noise is minor movement, a time window including walking vibrations can be extracted, making it possible to explore the internal structure of the ground or a structure.

[0050] As described above, vibrations originating from human activities, such as walking vibrations, are easily attenuated and are difficult to observe unless they occur near the sensor 3. Therefore, in a segment (time window) in which vibrations originating from human activities are determined to be predominant, it is considered that specific human activities, such as walking, are taking place near the sensor 3, not far from it. Since the vibrations are generated near the sensor 3, the vibrations that return to the upper surface 2 as reflected waves are predominantly vertical vibrations. In such cases, it is known that the S / N ratio of the reflected waves obtained by autocorrelation analysis is improved. In other words, if the exploration method described above is used, vertical vibrations will also predominate, and the S / N ratio of the measured reflected waves can be improved.

[0051] Although the embodiments of the present invention and modifications based thereon have been described above, the present invention is not necessarily limited to these examples. In addition, a person skilled in the art will be able to find various alternative embodiments and modifications without departing from the spirit of the present invention or the scope of the appended claims. [Explanation of symbols]

[0052] 1 structure 2 Top surface 3 Sensors 4 Arithmetic section 10. Internal Structure Exploration System

Claims

1. A method for exploring an internal structure of a structure having an upper surface to which walking vibration is constantly applied, comprising the steps of: A method for exploring the internal structure of a structure, comprising the steps of: attaching a sensor to the structure to measure micro-vibrations, which are changes in vibration over time; extracting a time window in which the walking vibration is imparted from the micro-vibrations; and performing an autocorrelation analysis on the vertical vibration contained in the walking vibration that is input into the interior of the structure from the upper surface and reflected therefrom, thereby exploring the internal structure of the structure.

2. 2. The method for exploring the internal structure of a structure according to claim 1, further comprising a step of integrating the vertical vibrations to separate signals.

3. A frequency Fp having a maximum power spectrum and a frequency Fs having a second maximum power spectrum in a frequency spectrum of an envelope of a waveform obtained by filtering the measured noise vibration are obtained, 2. The method for exploring the internal structure of a structure according to claim 1, wherein the time window is extracted when the frequency Fs is a harmonic of the frequency Fp and the power spectrum of the frequency Fp is greater than or equal to a predetermined value compared to the power spectra of other frequencies.

4. 2. The method for exploring the internal structure of a structure according to claim 1, wherein the extraction of the time window is performed based on a signal from a detection means for detecting the entrance and exit of a pedestrian onto and from the upper surface.

5. 2. The method for exploring the internal structure of a structure according to claim 1, wherein the structure is a ground with a paved surface.

6. 2. The method for examining the internal structure of a structure according to claim 1, wherein the structure is a bridge.

7. 7. The method for exploring the internal structure of a structure according to claim 1, wherein the sensor is a speed sensor or an acceleration sensor.

8. A system for detecting an internal structure of a structure having an upper surface to which walking vibration is constantly applied, comprising: A sensor for measuring noise vibrations, which are changes in vibration over time applied to the structure; a calculation unit that searches for an internal structure of the structure from the noise and vibrations, The calculation unit extracts a time window in which the walking vibration is imparted from the noise and vibration, and then performs an autocorrelation analysis on the vertical vibration contained in the walking vibration that is input into the interior of the structure from the upper surface and reflected.

9. 9. The system for investigating the internal structure of a structure according to claim 8, wherein the calculation unit performs a process including integrating the vertical vibration and separating the signals.

10. A frequency Fp having a maximum power spectrum and a frequency Fs having a second maximum power spectrum in a frequency spectrum of an envelope of a waveform obtained by filtering the measured noise vibration are obtained, 9. The internal structure exploration system of a structure as described in claim 8, characterized in that the time window is extracted when the frequency Fs is a harmonic of the frequency Fp and the power spectrum of the frequency Fp is larger than the power spectrum of other frequencies by a predetermined amount.

11. a detection means for detecting the ingress and egress of pedestrians onto and from said upper surface; 9. The system for investigating the internal structure of a structure according to claim 8, wherein the calculation unit extracts the time window based on a signal from the detection means.

12. The system for detecting the internal structure of a structure according to claim 8, wherein the structure is a ground surface with a paved surface.

13. 9. The system for detecting the internal structure of a structure according to claim 8, wherein the structure is a bridge.

14. 14. The system for detecting an internal structure of a structure according to claim 8, wherein the sensor is a speed sensor or an acceleration sensor.

Images