JP4728730B2 - Building health diagnosis method and health diagnosis program based on microtremor measurement - Google Patents

Description

  The present invention relates to a soundness diagnosis method and soundness diagnosis program for buildings based on microtremor measurement. More specifically, the present invention relates to a method and a program for determining the presence or absence of damage to a building due to an earthquake, strong wind, or the like based on measurement of minute vibration called microtremor. In the present invention, the soundness of a building refers to the soundness related to the structure of the building, such as the presence or absence of structural damage.

  As a conventional method for diagnosing soundness based on measurement of building vibration, there is, for example, a building soundness diagnosis method based on microtremor measurement (Patent Document 1).

  As shown in FIG. 20, this building health diagnostic method measures the microtremors of the building at the time of health and evaluation (S101, S101 '), and the cross spectrum and power at the time of health and evaluation from the measurement records. The spectrum is calculated (S102, S103, S103 ′), and the vibration characteristics of the building are calculated by obtaining the natural frequency and the natural mode from the calculation results of the spectra (S104, S105, S105 ′). Then, by comparing the calculation results of the vibration characteristics at the time of soundness and evaluation (S106), the soundness of the whole building is judged as good (S107). When it is determined that the soundness of the building is lost, the rigidity distribution of the building is calculated from the calculation result of the vibration characteristics (S108, S109, S109 ′), and the calculation result of the rigidity distribution at the time of healthy and evaluation is obtained. By comparing (S110), a position inferior to soundness and its degree are determined (S111).

  Moreover, as a method for diagnosing soundness by paying attention to vibration information of the building before and after an event that affects the soundness of the building, there is a structural performance index estimation device and a structure performance real-time monitoring method (Patent Document 2). ).

  In this structural performance real-time monitoring method, as shown in FIG. 21, the measurement points for measuring the absolute acceleration output when the basic acceleration is input are set in each layer of the structure and then measured at the measurement points. The apparatus is installed (first step), and the absolute acceleration in each layer of the structure observed through the measuring device is transmitted to the structural performance index estimating apparatus body (second process). Using the absolute acceleration for each layer, the damping coefficient and the rigidity are estimated for each layer of the structure by the arithmetic expression stored in the structural performance index estimating apparatus main body (third step). Then, the second step and the third step are repeated to obtain the estimated values related to the damping coefficient and the stiffness in each layer of the structure in time series, and the structural performance is monitored (fourth step).

  In other words, the method of Patent Document 2 evaluates vibration measurement data in real time, evaluates the rigidity and damping of each building layer (floor) from moment to moment, and detects these changes to detect the health of the building in real time. It is a method to evaluate with.

JP 2003-322585 A Japanese Patent Application Laid-Open No. 2005-083975

  In the method of Patent Document 1, it is assumed that the natural frequency and rigidity of the building are stable, and the natural frequency and the like are constant values if the soundness is the same without damage to the building. It is said. In other words, the natural frequency and stiffness change due to damage to the building, but do not change depending on other factors.Therefore, if there is a difference between the natural frequency and the like between the healthy state and the evaluation, the building will be damaged. It is assumed that it can be judged.

  However, in the examination of soundness diagnosis based on the vibration measurement of the building, the inventor does not stabilize the natural frequency of the building at a constant value, and fluctuates periodically throughout the day. It was found that the fluctuations are not regular fluctuations, depending on environmental conditions such as temperature, solar radiation, and wind power, and showing complex aspects. That is, it has been found that the natural frequency of the building fluctuates due to environmental conditions and the like even when the building is not damaged at all. Further, since the stiffness distribution of the building is an index estimated based on the natural frequency or the like, the apparent stiffness distribution also varies as the natural frequency varies. Therefore, under conventional assumptions, when comparing natural frequencies at different points in time, damage has occurred because the natural frequency has changed despite no damage to the building. Or because the natural frequency has not changed in spite of the fact that the building is damaged, it may be erroneously determined that no damage has occurred. .

  Furthermore, the method of Patent Document 1 is a method of diagnosing the soundness of a building by comparing the natural frequency and the like at the time of soundness and evaluation, and data indicating the state of soundness immediately after construction is required. If there is no data, the health of the building cannot be diagnosed. For this reason, there exists a problem that the object which can implement soundness evaluation will be restricted.

  In addition, since the behavior of buildings during earthquakes and strong winds is more complicated than the behavior at the time of microdeformation such as microtremors, in reality, in the real time as in the method of Patent Document 2 for the following reasons. It is considered difficult to estimate stiffness and damping.

  As shown in FIG. 22, the relationship of deformation δ-load P when a lateral load P is applied to a reinforced concrete single column on which a vertical load N is applied in the axial direction will be described as an example (Akinori Shibata: latest earthquake-resistant structure) Analysis, Morikita Publishing, 1981, p114).

  In FIG. 22, the rigidity is given by a value obtained by dividing the increase amount of the load P by the increase amount of the deformation δ, in other words, the tangential slope of the deformation δ-load P curve.

  Here, when a large deformation δ occurs as in an earthquake, the deformation δ-load P characteristic becomes a spindle-shaped curve, and the deformation δ and the load P are in a non-linear relationship. The stiffness will change from moment to moment depending on. In addition, since an actual building is composed of a plurality of columns, beams, and walls, the rigidity of the entire building exhibits a more complicated behavior than a single column. Therefore, the rigidity in a state where the building is damaged exhibits a more complicated behavior than the case of a single pillar as shown in FIG.

  From the viewpoint of stiffness estimation, the instantaneous stiffness that changes from moment to moment is essentially a short-time vibration around the time “near the time at which the instantaneous stiffness affects”. It shows that it must be estimated based on measurement data. However, as the vibration measurement data becomes shorter, the amount of information related to rigidity decreases, and the influence of accidental estimation errors increases, so the instantaneous rigidity is estimated from the short-time vibration measurement data. There is a problem that there is a limit.

  On the other hand, in order to solve the above problem, if the length of the vibration data is increased in order to improve the accuracy of the instantaneous stiffness, data of a time zone in which the instantaneous value of the stiffness is different is included. Therefore, as a result, the accuracy of instantaneous rigidity is lowered.

  From the above, it can be said that it is actually difficult to track in real time a complicated change in rigidity in a state where a large deformation occurs in the building as in the case of an earthquake as in the method of Patent Document 2.

  Therefore, the present invention is capable of diagnosing soundness without being affected by periodic fluctuations in the building health diagnosis index itself, and by utilizing a deformation-load relationship in a linear range. An object of the present invention is to provide a method for diagnosing the health of a building capable of stably diagnosing the health.

In FIG. 22, the single column deformation δ and the load P at the time of fine movement are both small behaviors as compared to the large deformation δ as in the case of an earthquake, and change near the origin (range within the frame 101. In the case of fine movement, the horizontal axis deformation δ is a very small range of 10 ⁻⁴ cm or less, and the frame 101 does not show an accurate range). In this region, the deformation δ-load P relationship is almost a straight line, and the deformation δ and the load P have a substantially linear relationship, and its rigidity (the slope of the deformation δ-load P relationship) is a constant value.

  From this, it is easy to infer that the deformation-load relationship of an actual building that combines multiple pillars, beams, and walls is a linear shape, and therefore the building rigidity must also be a stable and constant value. It becomes. From the perspective of stiffness estimation, the building stiffness remains stable in both the healthy and evaluation states, and the value does not change from moment to moment, so the measurement time for vibration data is constant. By ensuring the time, it is possible to improve the rigidity estimation accuracy.

  In the above, rigidity is taken as an example, and it is stated that the estimation accuracy of rigidity is better when estimating stable rigidity at the time of microdeformation such as at the time of fine movement rather than estimating instantaneous rigidity at the time of damage. It was. Here, the natural frequency is an amount determined by the rigidity and mass of the building, and considering that the mass becomes a constant value regardless of the damage of the building, the above discussion on rigidity remains unchanged for the natural frequency. apply. In other words, even if the natural frequency is used as an index for building health diagnosis, the stable natural frequency at the time of minute deformation such as at the time of fine movement is always estimated rather than the instantaneous amount estimated in real time. The estimation accuracy can be improved by estimating using vibration data of a certain length.

Therefore, based on the new knowledge unique to the inventor and utilizing a stable relationship between deformation and load of the building, the building health diagnosis method based on microtremor measurement according to claim 1 Collecting microtremor measurement data by always performing microtremor measurement and extracting microtremor measurement data before and after the occurrence of an event that may affect the health of the building. Each of the measurement data is divided into a plurality, and an index for diagnosing the soundness of the building is calculated for each divided microtremor measurement data, and the time y of the microtremor measurement data divided from the index y before the event occurs The constants ki and mi of the regression line (Equation 1) and the index y after the occurrence of the event and the regression line (Equation 1) of the regression line (Equation 1) with the time x of the divided microtremor measurement data are estimated, That By comparing the sections mi Les so that to diagnose the health of the building.
(Equation 1) yi = ki × xi + mi
Here, k: slope of regression line, subscript i: i = 1 for pre-event data, i = 2 for post-event data.

The building health diagnosis program based on the microtremor measurement according to claim 2 uses a computer for diagnosing the health of the building as a microtremor of the building before and after the occurrence of an event that may affect the health of the building. The means to input the measurement data, the means to divide the microtremor measurement data before and after the occurrence of the event into multiple parts, and calculate the index for diagnosing the health of the building for each divided microtremor measurement data Means, constants ki and mi of the regression line (Equation 2) between the index y before the occurrence of the event and the time x of the microtremor measurement data divided, and the time x of the microtremor measurement data divided from the index y after the occurrence of the event And a means for estimating constants ki and mi of the regression line (Equation 2) and a means for diagnosing the soundness of the building by comparing each intercept mi of the regression line.
(Equation 2) yi = ki × xi + mi
Here, k: slope of regression line, subscript i: i = 1 for pre-event data, i = 2 for post-event data.

Furthermore, the building health diagnosis program based on the microtremor measurement according to claim 3 is a computer program for diagnosing the health of the building, and the microtremor measurement data collected by constantly measuring the microtremor of the building is collected. Means to input, means to extract microtremor measurement data before and after the occurrence of an event that can affect the soundness of the building from microtremor measurement data, and microtremor measurement data before and after the event Means for dividing into a plurality of means, means for calculating an index for diagnosing the soundness of a building for each divided microtremor measurement data, a regression line between the index y before the occurrence of the event and the time x of the microtremor measurement data divided ( It means for estimating the constants ki and mi the regression line (equation 3) with constant ki and mi and time x of the measurement data microtremor divided as an index y after event occurrence in equation 3), regression By comparing the respective sections mi line it is caused to function as means for diagnosing the soundness of the building.
(Expression 3) yi = ki × xi + mi
Here, k: slope of regression line, subscript i: i = 1 for pre-event data, i = 2 for post-event data.

  Therefore, according to this building health diagnosis method and health diagnosis program, the index itself for diagnosing building health such as natural frequency fluctuates periodically under the influence of environmental conditions such as temperature, solar radiation, and wind power. Even if it is a building that does, the health can be diagnosed without being affected by the fluctuation. Further, since the index is calculated based on the microtremor data at all times, the soundness can be stably diagnosed on the premise of the deformation-load relationship in the linear range. In addition, since it does not require measurement data at the time of completion of the building or immediately after that, or an index of soundness based on it, it is possible to diagnose the soundness of a building regardless of whether it is a new building or a building that has passed a considerable amount of time. it can.

  As described above, according to the soundness diagnosis method and soundness diagnosis program of a building of the present invention, soundness diagnosis can be performed without being affected by periodic fluctuations in the soundness diagnosis index itself such as the natural frequency. In addition, the soundness diagnosis can be performed on the premise of the deformation-load relationship in the linear range, so that the reliability of the soundness diagnosis can be improved. Furthermore, since the soundness diagnosis of a building can be performed regardless of whether it is a new building or a building after a considerable period of time, it is possible to deal with various uses.

  Hereinafter, the configuration of the present invention will be described in detail based on the best mode shown in the drawings.

  FIG. 1 to FIG. 10 show an example of an embodiment of a building health diagnostic method based on microtremor measurement of the present invention.

  As shown in the flow chart of FIG. 1, this building health diagnostic method collects measurement data of microtremors by constantly measuring microtremors of buildings and before and after the occurrence of events that may affect the health of the buildings. The microtremor measurement data is extracted (S1), the microtremor measurement data before and after the event occurrence is divided into a plurality of pieces (S2), and the soundness of the building is divided for each divided microtremor measurement data. (S3), the regression line of the index before the event occurrence and the regression line of the index after the event occurrence are estimated (S4), and the health of the building is compared by comparing each intercept of the regression line Diagnose (S5).

  In applying the soundness diagnosis method of the present invention, first, the measurement of the microtremors of the building that is constantly collected and collected from the microtremor measurement data that is constantly collected, the events that may cause structural damage to the building such as earthquakes and strong winds. The microtremor data of the building before and after the occurrence are extracted (S1). Here, the constant fine movement refers to minute vibrations excited by, for example, daily winds or traffic vibrations. In the following, an event that can cause structural damage to a building such as an earthquake or a strong wind is called an event.

  Measurement of microtremor is always performed using, for example, a vibration sensor installed in a building. The extraction of microtremor data from the microtremor measurement data of a building that is constantly measured uses, for example, a method that combines a pretrigger measurement method and a level trigger method that have been used for earthquake observations of the ground and buildings. Do it.

  As shown in FIG. 2, the combination of the pre-trigger measurement method and the level trigger method is as follows. First, for example, the amplitude excess on the plus side and the minus side exceeds the center of the amplitude of vibration measured by a vibration sensor or the like. Threshold values are set as trigger levels TL1 and TL2.

  The trigger levels TL1 and TL2 take into consideration the levels of daily wind, traffic vibration, and other artificial vibrations, and the operator sets an appropriate threshold value so that the threshold value is not exceeded for vibrations other than the event targeted by the present invention. Set. Specifically, for example, it can be considered to be about 1 and −1 [gal], but the trigger levels TL1 and TL2 are not limited to this, and may be larger or smaller than this.

  In the method combining the pre-trigger measurement method and the level trigger method, when the vibration amplitude of the measured vibration waveform 1 exceeds the trigger level TL1 or TL2 (trigger point TP1 in FIG. 2), that point is the event occurrence point. The measurement time T1 before the event occurrence time T0 and the measurement time T2 after the event occurrence time T0 are set as the measurement target time, and the vibration data within the measurement target time is set as the final measurement target R1.

  The measurement time T1 is set by the operator in consideration of the amount of data (time) required for calculation of an index for the soundness diagnosis of the building before the event described later. Also, the measurement time T2 is set by the operator in consideration of the duration of the event itself and the amount of data (time) required for calculation of an index for building health diagnosis after the event occurs. Specifically, for example, each of the measurement times T1 and T2 is 5 minutes or more, preferably 10 minutes or more, more preferably 20 minutes or more, still more preferably 30 minutes or more, and even more preferably 40 minutes or more. Note that the measurement times T1 and T2 may be the same length or different lengths.

  Note that the method of determining the measurement target time is not limited to the pre-trigger measurement method, but a method that can determine the measurement target time of microtremor data required for calculating the index for building health diagnosis before and after the event occurs. Any method can be used. For example, a pre / post-trigger measurement method may be used instead of the pre-trigger measurement method as a method for determining the measurement target time.

  In the pre-post trigger measurement method, as shown in FIG. 3, when the vibration amplitude of the measured vibration waveform 1 exceeds the trigger level TL1 or TL2 (trigger point TP1 in FIG. 3), an event is generated at that point. The process is the same as the pre-trigger measurement method until the time T0 and the measurement time T1 before the event occurrence time T0 is set as the measurement target time before the event occurrence.

  In the pre-post-trigger measurement method, when the vibration amplitude of the vibration waveform 1 corresponding to the event becomes smaller than the trigger levels TL1 and TL2 (post-trigger point TP2), that point is set as the event end point T3, and the event occurrence point T0. Measurement time T4 from the event end time T3 to the measurement target time during the event continuation, and measurement time T5 after the event end time T3 as the measurement target time after the event end. And by the above, measurement time T1, T4, and T5 are made into measurement object time, and the vibration data in the measurement object time are made into final measurement object R2.

  Unlike the pre-trigger measurement method, the pre-post trigger measurement method can automatically control the measurement time T4 that is the event duration. Therefore, when the pre / post-trigger measurement method is used, the measurement time T1 is determined by the operator in consideration of the amount of data (time) required for calculating an index for building soundness diagnosis before the event described later. The measurement time T5 is set by the operator in consideration of the amount of data (time) required for calculating the index for the soundness diagnosis of the building after the event occurs. Note that the measurement time T4, which is the event duration, varies depending on the type of event, but if the event is an earthquake, for example, it varies depending on the occurrence mechanism and propagation path of the earthquake, and ranges from less than 10 seconds to several tens of minutes depending on the earthquake motion. Have Therefore, the pre / post-trigger measurement method that can automatically control the event duration has a feature of ensuring a constant measurement time T5 after the occurrence of an event, regardless of the event duration, compared to the pre-trigger measurement method.

  In addition, as a method for extracting the microtremor data of the measurement objects R1 and R2 by the above-described method from the microtremor measurement data collected by constantly measuring the microtremors of the building, for example, the following two methods are used. Can be considered.

  In the first method, a recording device is provided in a vibration sensor or the like, and the latest continuous fine movement data corresponding to at least the measurement time T1 is recorded in the recording device while updating is repeated in normal times. Then, when the occurrence of an event is detected, the latest continuous fine movement data corresponding to the measurement time T1 is not updated and kept as data before the occurrence of the event. Further, the measurement time T2 after the occurrence of the event or This is a method for recording constantly fine movement data corresponding to T4 and T5 as data after the occurrence of an event. As a result, the continuous fine movement data of the measurement objects R1 and R2 are extracted from the continuous fine movement measurement data of the building that is constantly measured.

  The second method is to provide a recording device in the vibration sensor, etc., and record all the fine movement measurement data of the building that is constantly measured as magnetic data or paper data as it is, This is a method for extracting fine movement data of the measurement objects R1 and R2 from the data.

  Here, the recording device is, for example, a magnetic disk drive and a magnetic disk that can be inserted and removed. In addition, the update is, for example, recording by overwriting new data on old data recorded on a magnetic disk.

  The detection method of event occurrence is not limited to the level trigger method, and for example, the STA / LTA method may be used.

  The STA / LTA method is a method of taking an average vibration amplitude ratio in two different time intervals and determining that an event has occurred when this ratio exceeds a preset threshold (STA: Short Time Average). , LTA: Long Time Average). Specifically, the average of the vibration amplitude in the short time interval is STA, the average of the vibration amplitude in the long time interval including the short time interval is LTA, and the STA / LTA value exceeds a preset threshold. It is determined that an event has occurred. For example, in the event of an earthquake, this is a method that uses the principle that the STA / LTA value increases because the average STA of vibration amplitude in a short time section increases rapidly compared to the average LTA of vibration amplitude in a long time section. .

  When the STA / LTA method is applied to the present invention, the time length of the short time section and the long time section and the threshold value of the STA / LTA are not particularly limited, and each worker is appropriate in consideration of the target event. Set the value. Specifically, for example, it is conceivable that a short time interval corresponding to STA is 1 to 2 seconds, a long time interval corresponding to LTA is 10 to 20 seconds, and a threshold of STA / LTA is about 3 to 5. .

  Next, the index of the soundness diagnosis of a building is calculated using the microtremor data obtained in S1 (S2 to S3). The index here refers to the natural frequency or rigidity of the building.

  In the present embodiment, the fine movement is always extracted by a method combining the pre-trigger measurement method and the level trigger method (S1), and as shown in FIG. 4, the vibration amplitude of the vibration waveform 1 is the trigger level TL1 at the trigger point TP1. A case will be described in which the constant fine movement data D1 and D2 in the measurement target R1 are obtained that exceed the measurement time T1 and T2 before and after the time T0 and have the measurement target time.

  First, the fine movement data D1 and D2 before and after the event occurrence in the measurement target R1 obtained as a result of S1 are respectively divided (S2).

  In the present embodiment, the continuous fine movement data D1 before the occurrence of the event is divided into three in order of time, and the continuous fine movement data D2 after the occurrence of the event is divided into four in order of time. In the soundness diagnosis method of the present invention, data corresponding to the vibration amplitude during the event is not used, and the continuous fine motion of the section including the vibration amplitude during the event in the quarterly fine motion data D2 after the event occurrence divided into four is used. Diagnosis is performed except for data d0. In other words, with respect to the constantly fine movement data D2 after the occurrence of the event, the remaining part excluding the data during the event is divided into three.

  As described above, in the present embodiment, the three divided data d1, d2, and d3 (hereinafter referred to as divided data d1 to d3) before the occurrence of the event and the three divided data d4, d5, and d6 (hereinafter referred to as the divided data) after the occurrence of the event. Diagnosis is performed using the divided data d4 to d6 and the divided data d1 to d6 together with the previous divided data d1 to d3.

  Note that the number of divisions of the fine movement data D1 and D2 is not limited to three divisions or four divisions, and when the fine movement measurement is performed using the pre-trigger measurement method, For the fine movement data D2 after the occurrence of the event, at least in two or more divisions, the remaining portion excluding data including the vibration amplitude during the event may be divided into at least two or more.

  The number of divisions of the microtremor data D1 and D2 is determined by the operator in consideration of the length of each of the measurement times T1 and T2 and the amount of data required for calculating the index for building health diagnosis. to decide. Further, the division number of the fine movement data D1 before the event occurrence and the division number of the fine movement data D2 after the occurrence of the event may be the same or different. Furthermore, the fine movement data D1 and D2 may be divided equally or evenly.

  Here, when the fine movement data is measured using the pre-post trigger measurement method, the fine movement data corresponding to the measurement times T1 and T5 is divided and used for calculation of the index. In this case, since the vibration data during the event continuation is not included in the continuous fine movement data corresponding to the measurement time T5, all the divided data are used without removing any of the divided data. Therefore, it is only necessary to divide the fine movement data before and after the occurrence of the event into at least two parts.

  Subsequently, an index for building health diagnosis is calculated for each of the divided data d1 to d6 divided in S2 (S3). Note that the method for calculating the natural frequency and rigidity of the building based on the microtremor data is a well-known method, and therefore details thereof are omitted here (for example, Patent Document 1).

  Furthermore, for each of the divided data d1 to d6, an intermediate time in the time segment of each divided data d1 to d6 is given as a time corresponding to each divided data d1 to d6.

  As described above, the data of the combination of the natural frequency of the building, the rigidity, and the time, which are indexes for soundness diagnosis, are arranged for each of the divided data d1 to d6.

  In this embodiment, the combination of the soundness diagnosis index and time for each of the divided data d1 to d3 obtained by dividing the microtremor data D1 before the occurrence of the event into three data in the order of time 2a, 2b and 2c (hereinafter, event occurrence) The data of the combination of the soundness diagnosis index and the time for each of the divided data d4 to d6 obtained by dividing the microtremor data D2 after the occurrence of the event into three in the order of time 3a, 3b and 3c (hereinafter referred to as post-event occurrence data 3a to 3c).

  Next, regression lines are estimated separately for the pre-event data 2a to 2c and the post-event data 3a to 3c (S4).

  Specifically, the constants ki and mi in (Expression 1) are estimated.

  (Formula 1) yi = ki × xi + mi

  Here, x: time, y: index of soundness diagnosis, k: slope of regression line, m: intercept of regression line, subscript i: i = 1 in case of pre-event data 2a to 2c, post-event data In the case of 3a-3c, i = 2.

  The regression line can be estimated using, for example, the least square method.

  Thereafter, when it can be determined that the soundness diagnosis index has not changed before and after the event occurrence (FIGS. 5 to 7) and when it can be determined that it has changed (FIGS. 8 to 10). Each case will be described. Specifically, when the pre-event occurrence data 2a to 2c and the post-event data 3a to 3c obtained as a result of S3 are plotted, for example, a case as shown in FIG. 5 and a case as shown in FIG. Will be described. 5 to 10, the horizontal axis represents time, the vertical axis represents a soundness diagnosis index, and time T0 in the figure represents an event occurrence time point.

  As shown in FIGS. 6 and 9, the regression line 4 of the pre-event data 2a to 2c and the regression line 5 of the post-event data 3a to 3c are estimated.

  Here, in the present invention, when estimating the regression lines 4 and 5, it is assumed that the slopes k1 and k2 of the regression lines 4 and 5 are the same. This is based on the knowledge that the natural frequency of the building fluctuates periodically, and it is assumed that the health diagnosis index fluctuates in the same way before and after the event. It is.

  The estimation of the regression lines 4 and 5 assuming that the slopes k1 and k2 are the same is performed, for example, by first giving the slope k1 obtained by estimating the regression line 4 as the slope of the regression line 5, Alternatively, the regression line 4 and the regression line 5 may be estimated, and the average values of the respective slopes k1 and k2 may be newly given as slopes to estimate the regression line 4 and the regression line 5. good.

  Subsequently, the soundness of the building is diagnosed (S5).

  The soundness of the building is diagnosed by comparing the intercepts m1 and m2 of the regression line 4 and the regression line 5 estimated in S4.

  Specifically, as a result of estimating the regression line 4 and the regression line 5, as shown in FIG. 6, when the intercepts m1 and m2 are the same or almost the same, the health diagnosis index is an event. It has not changed before and after the occurrence, and it is judged that the soundness of the building has not been impaired.

  This is based on the idea that, from the estimation results of the regression line 4 and the regression line 5, for example, a periodic fluctuation waveform 6 of an index as shown in FIG. 7 is assumed. That is, from FIG. 5 and FIG. 6, the soundness diagnosis index apparently changes over time. This apparent change is, for example, the soundness diagnosis index shown as the periodic fluctuation waveform 6 of the index. It is based on the knowledge that it is due to its own periodic fluctuations and that the index of health diagnosis is not essentially changing.

  Here, the magnitude of the difference between the intercepts m1 and m2 of the regression lines 4 and 5 for judging whether or not the soundness of the building is impaired depends on the size and variation of the calculated soundness diagnosis index. In consideration, an operator determines an appropriate reference value for each building to be diagnosed. For example, it is conceivable that the difference between the intercepts m1 and m2 is about 0.02 [Hz], or the change rate is about 1 [%]. However, the reference value of the difference or rate of change in diagnosing the soundness of the building is not limited to this, and a reference value smaller than this or a reference value larger than this may be used. good.

  In addition, when the index of the health diagnosis before and after the occurrence of the event is simply compared, specifically, for example, the index of the pre-event data 2c and the post-event data 3a are directly compared, or the pre-event data 2a to 2 When the average of each index of 2c and post-event data 3a to 3c is compared, an apparent difference between before and after the occurrence of the event even though the health diagnosis index is essentially unchanged. It can be seen that there is a possibility of making a wrong judgment that the soundness of the building is impaired due to

  On the other hand, as a result of estimating the regression line 4 and the regression line 5, as shown in FIG. 9, when the intercepts m1 and m2 are different from each other, the health diagnosis index changes before and after the occurrence of the event. Judge that the soundness of the

  This is based on the idea that, based on the estimation results of the regression line 4 and the regression line 5, for example, periodic fluctuation waveforms 6a and 6b having indices as shown in FIG. FIG. 10 shows that the index of health diagnosis essentially changes due to the occurrence of an event, and the period fluctuation waveform 6a of the index before the event occurrence changes to the period fluctuation waveform 6b of the index after the event occurrence at the boundary of the event occurrence time T0. It shows that.

  Here, the criteria for determining that the soundness of the building is impaired based on the difference Δm between the intercepts m1 and m2 of the regression line 4 and the regression line 5 are, for example, the intercepts m1 and m2 as described above. It is conceivable to use about 0.02 [Hz] as a reference for the difference Δm, or to use about 1 [%] for the rate of change. However, the reference value of the difference or rate of change in diagnosing the soundness of the building is not limited to this, and a reference value smaller than this or a reference value larger than this may be used. good.

  In addition, when simply comparing the health diagnosis index before and after the event occurrence, specifically, for example, when comparing the health diagnosis index of the pre-event data 2c and the post-event data 3a as they are, In spite of the change in the health diagnosis index, there is no apparent difference between before and after the event, so there is a possibility of misjudging that the health of the building is not impaired. I understand.

  By adopting the above procedure, even if the natural frequency and rigidity of the building fluctuate periodically depending on the temperature, solar radiation, wind power, etc. It becomes possible to evaluate the natural frequency and the amount of change in rigidity depending only on structural damage, and based on this, it becomes possible to diagnose the soundness of the building.

  Then, an example in the case of performing the soundness diagnostic method of the building of this embodiment using a program is shown.

  FIG. 11 shows the overall configuration of the health diagnostic apparatus 7 for executing the health diagnostic program. In this embodiment, a recording device is provided in the vibration sensor or the like, and all the measurement data of the microtremor of the building that is constantly measured is recorded as it is, and the pretrigger measurement is performed from the recorded microtremor data for a certain period. A case will be described in which continuous fine movement data of the measurement target R1 is extracted by a method combining a method and a level trigger method.

  A storage unit 9, an instruction input unit 10, a display unit 11, and a data input unit 12 are connected to the control unit 8. The memory 13 is built in or connected to the control unit 8.

  The control unit 8 performs the calculation related to the control of the whole health diagnosis device 7 and the soundness diagnosis by the soundness diagnosis program stored in the storage unit 9, and always collects the fine tremors of the building. The continuous fine movement data extraction processing section that extracts the continuous fine movement data of the building before and after the occurrence of the event from the continuous fine movement measurement data, and the continuous fine movement data division processing section that divides the continuous fine movement data before and after the occurrence of the event. , Health diagnostic index calculation processing unit that calculates building health diagnostic index for each divided data, regression line estimation processing unit that estimates regression line by data before event occurrence and data after event occurrence, intercept of calculated regression line The soundness determination processing unit for diagnosing the soundness of the building based on the value of is configured.

  The control unit 8 is a CPU, for example. The storage unit 9 is a hard disk, for example. The instruction input unit 10 is an interface for giving at least an operator command to the CPU, and is, for example, a keyboard. The display unit 11 is a display, for example. The data input unit 12 supplies at least data recorded on an electronic medium to the CPU, and is a magnetic disk drive, for example.

  First, the continuous fine movement data of the building before and after the occurrence of the event is extracted from the continuous fine movement measurement data collected by constantly measuring the fine movement of the building (S1). Here, the measurement and collection of microtremor data is always performed by a vibration sensor or the like provided with a recording device. All of the microtremor data before and after the event or all the microtremor data that is constantly measured is, for example, a magnetic disk drive or attached / detached. It is stored in a recording device such as a magnetic disk.

  The microtremor data extraction processing unit of the control unit 8 reads the time and vibration amplitude data corresponding to the time through the data input unit 12 as the microtremor measurement data recorded on the magnetic disk for a certain period.

  When the read vibration amplitude exceeds the trigger level TL1 or TL2, the time is stored in the storage unit 9 or the memory 13 as the event occurrence time T0. Then, the time corresponding to the measurement time T1 before the event occurrence time T0 and the vibration amplitude data corresponding to the time are read from the magnetic disk via the data input unit 12, and are always stored in the storage unit 9 or the memory 13 as the fine movement data D1. Remember. Further, the time corresponding to the measurement time T2 after the event occurrence time T0 and the vibration amplitude data corresponding to the time are read from the magnetic disk via the data input unit 12, and are always stored in the storage unit 9 or the memory 13 as the fine movement data D2. Remember.

  Next, the fine movement data D1 and D2 before and after the event occurrence obtained as a result of S1 are respectively divided (S2).

  The fine movement data division processing unit of the control unit 8 first reads the fine movement data D1 (time and vibration amplitude data corresponding to the time) stored in the storage unit 9 or the memory 13 in S1.

  Then, based on the time of the read fine movement data D1, the entire fine movement data D1 is divided into three parts in the order of the time and stored in the storage unit 9 or the memory 13 as divided data d1 to d3.

  Next, the fine movement data division processing unit of the control unit 8 reads the fine movement data D2 stored in the storage unit 9 or the memory 13 in S1.

  Then, based on the time of the read fine movement data D2, the entire fine movement data D2 is divided into three parts in the order of the time and stored in the storage unit 9 or the memory 13 as divided data d4 to d6.

  When the health diagnosis device 7 does not extract the constantly fine movement data in S1, instead of reading the fine movement data D1 and D2 stored in the storage unit 9 or the memory 13, the continuous fine movement data recorded on the magnetic disk is stored. The fine movement data D1 and D2 may be read via the data input unit 12.

  Next, an index for building health diagnosis is calculated for each of the divided data d1 to d6 divided in S2 (S3).

  First, the soundness diagnostic index calculation processing unit of the control unit 8 reads the divided data d1 stored in the storage unit 9 or the memory 13 in S2.

  Then, using the divided data d1 (time and vibration amplitude data corresponding to the time, which is data for a certain period of time), the natural frequency and rigidity of the building, which are indices for building health diagnosis, are calculated. .

  Further, based on the time information of the divided data d1, the intermediate time of the time division of the divided data d1 is calculated.

  Then, the calculated health diagnosis index value and intermediate time are stored in the storage unit 9 or the memory 13 as pre-event data 2a corresponding to the divided data d1.

  Similarly, the soundness diagnostic index calculation processing unit of the control unit 8 calculates the index and the intermediate time for the divided data d2 and d3, and the pre-event data 2b and 2c corresponding to the divided data d2 and d3. As a result, the value and time of each index are stored in the recording unit 9 or the memory 13.

  Similarly, the soundness diagnostic index calculation processing unit of the control unit 8 calculates the index and the intermediate time for the divided data d4 to d6, and the post-event data 3a to 3d corresponding to the divided data d4 to d6. The value of each index and the time are stored in the recording unit 9 or the memory 13 as 3c.

  Next, regression lines are estimated separately for the pre-event data 2a to 2c and the post-event data 3a to 3c (S4).

  The regression line estimation processing unit of the control unit 8 first reads the index values and times of the pre-event data 2a to 2c stored in the storage unit 9 or the memory 13 in S3, and based on this data, the slope of the regression line 4 is read. Estimate k1 and intercept m1.

  And it memorize | stores in the memory | storage part 9 or the memory 13 as inclination k1 and intercept m1 of the regression line 4 corresponding to the data 2a-2c before event occurrence.

  Similarly, the regression line estimation processing unit of the control unit 8 reads the index values and times of the post-event data 3a to 3c stored in the storage unit 9 or the memory 13 in S3, and based on this data, the regression line 5 Estimate the slope k2 and the intercept m2.

  And it memorize | stores in the memory | storage part 9 or the memory 13 as inclination k2 and intercept m2 of the regression line 5 corresponding to post-event generation data 3a-3c.

  Next, the soundness of the building is diagnosed using the intercepts m1 and m2 calculated in S4 (S5).

  The soundness determination processing unit of the control unit 8 reads the intercept m1 of the regression line 4 and the intercept m2 of the regression line 5 stored in the storage unit 9 or the memory 13 in S4.

  Subsequently, the soundness determination processing unit of the control unit 8 compares the values of the intercept m1 and the intercept m2. When it is determined that there is no difference in the values, it is determined that there is no difference in the health diagnosis index and that the building is not damaged. Then, as a diagnosis result of soundness, a diagnosis result indicating that there is no problem with soundness is stored in the storage unit 9 and displayed on the display unit 11 as necessary. On the other hand, when it is determined that there is a difference in value, it is determined that there is a difference in the index of the health diagnosis and the building is damaged. Then, as a diagnosis result of soundness, a diagnosis result indicating that there is a problem with soundness is stored in the storage unit 9 and displayed on the display unit 11 as necessary.

  In the above description, the control unit 8 receives instructions from the operator through the instruction input unit 10 as appropriate, and performs control such as data selection and processing start.

  In addition, although the above-mentioned form is an example of the suitable form of this invention, it is not limited to this, A various deformation | transformation implementation is possible in the range which does not deviate from the summary of this invention. For example, in the present embodiment, the case where both the pre-event data 2a to 2c and the post-event data 3a to 3c tend to increase gradually has been described. For example, the pre-event data 2a to 2c tends to increase gradually and after the event occurs. When the data 3a to 3c tend to gradually decrease, a gradual increasing tendency starting from the post-event data 3a having the same absolute value of the change and the gradual decreasing tendency to 3b and 3c starting from the post-event data 3a is assumed. Thus, the soundness diagnosis can be performed in the same manner as the method exemplified in the present embodiment.

  In addition, as an embodiment in the case of performing using a program, a case where a vibration sensor and the like and a soundness diagnosis device that measure microtremors are configured separately has been described as an example. The sex diagnostic apparatus may be configured as an integral unit.

  Further, in this embodiment, the case where the periodic fluctuation of the health diagnosis index itself is relatively large has been described. However, the periodic fluctuation of the health diagnosis index itself is very gradual and the slope of the regression line is not clear. In some cases, it is possible to perform a soundness diagnosis with a slope of zero.

  As an embodiment of the present invention, FIG. 12 to FIG. 19 show a central building of the Abiko District South Research Building (located in Abiko City, Chiba Prefecture, 10-story office building) (hereinafter referred to as a target building). The result of microtremor measurement and soundness diagnosis performed as a target is shown.

  The response amplitude observed over 30 days from February 1 to March 2, 2005, and the natural frequency in the north-south direction and east-west direction were arranged every 15 minutes in the target building. The results shown in FIGS. 13 and 14 were obtained.

  First, it was confirmed that the natural frequency fluctuates periodically in units of days from the results of arranging the temporal changes in FIGS. 13 and 14.

  In addition, an earthquake with the epicenter in the southern part of Ibaraki Prefecture occurred at 4:46 am on February 16 (at the time indicated by ▼ in FIGS. 12 to 14), and a shaking with a seismic intensity of about III was observed near the target building. It was.

  Here, in order to compare with the diagnosis result of the soundness diagnosis method of the present invention, the natural frequency is calculated using measurement data from 15 noon every 15 minutes, and the soundness of the building based on the change of the natural frequency every day is calculated. Attempted diagnosis. When it is assumed that microtremor data is measured to diagnose the soundness of buildings based on long-term changes in natural frequency, it is practically necessary to calculate the natural frequency once a day (for example, for calculating natural frequency). The measurement data from 15 noon to 15 minutes is used in consideration of the fact that it is possible to allow labor and cost if it is a necessary time) measurement.

  The changes over time in the natural frequency calculated using the measurement data from 15 noon to 15 minutes were organized, and the results shown in FIGS. 15 and 16 were obtained. In addition, the position of ▼ in the figure indicates the earthquake occurrence time.

  From this result, it is difficult to read the change in natural frequency before and after the occurrence of the earthquake because the natural frequency varies widely, and the occurrence of damage to the building due to the earthquake based on the daily change in natural frequency. It was confirmed that it was difficult to determine the presence or absence.

  Next, the soundness diagnosis was performed using the soundness diagnosis method for buildings according to the present invention.

  First, the data of the morning of February 16 when the earthquake occurred was divided into 15-minute data, and changes over time in response amplitude and natural frequency were organized. The results shown in FIGS. 17, 18 and 19 were obtained. Obtained. In addition, the position of ▼ in the figure indicates the earthquake occurrence time.

  In the present invention, for example, a combination of a pre-trigger measurement method and a level trigger method is actually measured as microtremor data for the primary natural frequency in the north-south direction shown in FIG. 18 and the primary natural frequency in the east-west direction shown in FIG. Assuming the range, we diagnosed the soundness using three divided data for each before and after the earthquake. The divided data including the vibration amplitude immediately after the earthquake occurred was excluded from the data used for diagnosis.

  For the natural frequency of the north-south direction primary shown in FIG. 18, a regression line was estimated using the three divided data before the occurrence of the earthquake. In this example, since the periodic fluctuation of the natural frequency itself was very gentle, the intercept was estimated with the slope of the regression line as zero. As a result, 1.758 [Hz] was obtained as an intercept of the regression line of the pre-earthquake data.

  Furthermore, the slope was set to zero as in the case of the regression line of the data before the earthquake occurrence, and the intercept of the regression line was estimated using the three divided data after the earthquake occurrence to obtain 1.719 [Hz].

  Moreover, the regression line was estimated using the three division | segmentation data before an earthquake occurrence about the natural frequency of the east-west primary frequency shown in FIG. As described above, since the periodic fluctuation of the natural frequency itself was very gradual, the intercept was calculated with the slope of the regression line set to zero. As a result, 2.002 [Hz] was obtained as an intercept of the regression line of the pre-earthquake data.

  Furthermore, the slope was set to zero in the same manner as the regression line of the data before the earthquake occurrence, and the intercept of the regression line was calculated using the three divided data after the earthquake occurrence to obtain 1.971 [Hz].

  From the above, 0.039 [Hz] was obtained in the north-south direction and 0.031 [Hz] in the east-west direction as the difference between the regression line intercepts of the pre-earthquake data and the post-earthquake data. The change rates were 2.2 [%] and 1.5 [%], respectively. From this result, it was determined that the target building was damaged by the earthquake that occurred on February 16th.

  From the above results, by using the soundness diagnosis method of the present invention, it is possible to clearly distinguish changes in natural frequency before and after an earthquake even when the natural frequency itself fluctuates periodically. It was confirmed that this is a method that can stably diagnose the soundness of buildings.

It is a flowchart explaining an example of embodiment of the soundness diagnostic method of the building based on the microtremor measurement of this invention. It is a schematic diagram explaining the pre-trigger measurement method. It is a schematic diagram explaining the pre-post trigger measurement method. It is a schematic diagram explaining constantly fine movement data of the embodiment. It is a figure which shows the data of the combination of the time of embodiment and the parameter | index of soundness diagnosis. It is a figure which shows the regression line of the data shown in FIG. It is a figure explaining the period fluctuation waveform of the parameter | index assumed when the regression line shown in FIG. 6 is estimated. It is a figure which shows the other data of the combination of the time of embodiment and the parameter | index of a soundness diagnosis. It is a figure which shows the regression line of the data shown in FIG. It is a figure explaining the period fluctuation waveform of the parameter | index assumed when the regression line shown in FIG. 9 is estimated. It is a figure which shows the whole structure of the soundness diagnostic apparatus for implementing the soundness diagnostic program of embodiment. It is a figure which shows the measurement data of the response amplitude of an Example. It is a figure which shows the natural frequency (north-south direction primary) of an Example. It is a figure which shows the natural frequency (east-west direction primary) of an Example. It is a figure which shows the natural frequency (the north-south direction primary, every noon) of an Example. It is a figure which shows the natural frequency (East-west direction primary, every noon) of an Example. It is a figure which shows the measurement data (divided into 15 minutes only in the morning) of the response amplitude of an Example. It is a figure which shows the natural frequency (The north-south direction primary, the morning only, and 15 minutes division | segmentation) of an Example. It is a figure which shows the natural frequency (East-west direction primary, only in the morning, divided into 15 minutes) of an Example. It is a flowchart explaining the soundness diagnostic method of the building based on the conventional microtremor measurement. It is a flowchart explaining the conventional structural performance parameter | index estimation apparatus and the structural performance real-time monitoring method of a structure. It is a figure explaining the displacement-load characteristic of a reinforced concrete column.

Explanation of symbols

2a to 2c Data before event occurrence 3a to 3c Data after event occurrence 4 Regression line of data before event occurrence 5 Regression line of data after event occurrence 7 Health diagnostic device

Claims (3)

Collecting the microtremor measurement data by always performing microtremor measurement of the building and extracting the microtremor measurement data before and after the occurrence of the event that may affect the soundness of the building. In addition, each of the microtremor measurement data after the occurrence of the event is divided into a plurality of pieces, and an index for diagnosing the soundness of the building is calculated for each of the divided microtremor measurement data. the index y and the divided constants ki and mi and the index y and time x of the measurement data of said divided microtremors after the event occurrence of the regression line of the time x of the measurement data microtremor (equation 1) estimate the constants ki and mi the regression line (equation 1) with, healthy building by comparing the respective sections mi of the regression line based on microtremor measurement, which comprises diagnosing the soundness of the building Sex diagnosis .
(Equation 1) yi = ki × xi + mi
Here, k: slope of regression line, subscript i: i = 1 for pre-event data, i = 2 for post-event data. A computer for diagnosing the health of the building, means for inputting measurement data of microtremors of the building before and after the occurrence of an event that may affect the health of the building, before the occurrence of the event, and after the occurrence of the event Means for dividing each of the microtremor measurement data into a plurality of means, means for calculating an index for diagnosing the soundness of the building for each of the divided microtremor measurement data, the index y before the occurrence of the event, and A regression line between the constants ki and mi in the regression line (Formula 2) of the divided microtremor measurement data with time x and the index y after the occurrence of the event and the time x of the segmented microtremor measurement data. It means for estimating the constants ki and mi (formula 2), the building based on microtremor measurement for comparing the respective sections mi of the regression line to function as a means of diagnosing the soundness of the building Ken Sex diagnostic program.
(Equation 2) yi = ki × xi + mi
Here, k: slope of regression line, subscript i: i = 1 for pre-event data, i = 2 for post-event data. A computer for diagnosing the soundness of the building, means for inputting the measurement data of the microtremor collected by constantly measuring the microtremor of the building, and influences the health of the building from the measurement data of the microtremor Means for extracting the measurement data of the microtremor before and after the occurrence of an event that can be given, means for dividing the measurement data of the microtremor before and after the occurrence of the event into a plurality of parts, and the divided microtremor Means for calculating an index for diagnosing the soundness of the building for each measurement data, a constant of a regression line (Formula 3) between the index y before the event occurrence and the time x of the divided microtremor measurement data ki and mi and means for estimating the constants ki and mi the regression line (equation 3) and the index y and time x of the measurement data of said divided microtremors after said event occurs, the regression line of its Microtremor health diagnostics building based on the measurement of the order to compare the sections mi of, respectively function as a means for diagnosing the soundness of the building.
(Expression 3) yi = ki × xi + mi
Here, k: slope of regression line, subscript i: i = 1 for pre-event data, i = 2 for post-event data.

Images