Method and apparatus for preventing serious disasters associated with the collapse of building resilience

Abstract

According to one embodiment of the present disclosure, there is provided a method for preventing serious disasters associated with the collapse of a building by predicting the stiffness of the building using information including vibration data collected by a vibration sensor 120 attached to the building, the method including: (a) a step in which vibration data and the stiffness of the building are collected by a learning information generation module 10; (b) a step in which a vibration characteristic learning module 20 learns a prediction model 22; (c1) a step in which the stiffness of the building is output when vibration data is input to the vibration characteristic learning module 20; (d) a step in which the stiffness output in the step (c1) is compared with the magnitude of a pre-set danger level; (e2) a step in which, if the output stiffness in the step (d1) is less than the pre-set danger level, a step in which a cumulative impact amount applied to the building during a predetermined unit period is compared with a resilience threshold of the building; and (f2) a step in which, if the cumulative impact amount in the step (e2) exceeds the resilience threshold of the building, an alarm providing module 40 provides an alarm to building users, an adjacent building management department, and an infrastructure management department.

Description

[Technical field]

[0001] The present disclosure relates to a system and method for preventing serious disasters caused by the collapse of buildings, and more particularly to a system and method for preventing serious disasters caused by destruction due to deterioration of buildings or destruction due to demolition of buildings. [Background technology]

[0002] The material described in this section is merely intended to provide background information for the present disclosure and may not constitute prior art.

[0003] In recent years, there have been frequent incidents of building collapses resulting in deaths, disappearances, and financial losses amounting to hundreds of billions of yen. Furthermore, many human casualties have been caused by the demolition of buildings and the collapse of retaining walls during redevelopment.

[0004] In order to minimize the damage associated with such building collapses, there has been an increase in research into predicting abnormal behavior of buildings and providing alarms.

[0005] However, previous studies have only used vibration sensors to monitor building behavior, resulting in low accuracy.

[0006] In addition, the system has a problem of low accuracy because it does not take into account impacts from building users or deterioration of buildings due to abnormal weather, natural disasters, and other disasters.

[0007] This makes it possible to further measure the rigidity of the building, but there is a problem in that it is difficult to perform precise measurements.

[0008] In addition, the huge amount of information cannot be systematically classified or pre-processed, resulting in low accuracy in predicting building collapses.

[0009] In addition, the rigidity of a building may temporarily weaken. In this case, there is a problem that a false disaster warning is given without taking into consideration that the rigidity of the building will be restored, causing confusion.

[0010] Furthermore, even though a building collapse would cause severe damage to adjacent buildings and infrastructure, there is a lack of a system that can systematically notify such damage. Summary of the Invention [Problem to be solved by the invention]

[0011] The present disclosure aims to provide a system and method capable of predicting the stiffness of a building with high accuracy using vibration data collected in real time by a vibration sensor.

[0012] In addition, the present disclosure aims to provide a system and method that can increase convenience by automatically collecting information on the congestion level of building users, abnormal weather, and disaster information without the user having to directly input it.

[0013] In addition, the present disclosure aims to provide a system and method that can obtain highly reliable results even under conditions where it is impossible to apply an ultrasonic flaw detection device or where the results of the ultrasonic flaw detection device are difficult to trust.

[0014] Another object of the present disclosure is to provide a system and method capable of obtaining high-quality learning data by removing outliers in input information.

[0015] It is also an object of the present disclosure to provide a system and method that can take into account various phenomena that affect concrete.

[0016] Another object of the present disclosure is to provide a system and method that can further improve learning accuracy by labeling input information.

[0017] It is also an object of the present disclosure to provide a system and method that can prevent false alarms by measuring and determining a composite resilience index of the stiffness of a building structure.

[0018] Another object of the present disclosure is to provide a system and method capable of minimizing human damage and secondary damage.

[0019] The problems to be solved by the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description. [Means for solving the problem]

[0020] According to an embodiment of the present disclosure, a method for preventing serious disasters due to the collapse of a building by predicting the stiffness of the building using information including vibration data collected by a vibration sensor 120 attached to the building, the method comprising the steps of: (a) collecting input information including vibration data and output information including the stiffness of the building by a learning information generation module 10; (b) a vibration characteristic learning module 20 learning a prediction model 22 using the input information and the output information; (c1) outputting the stiffness of the building when vibration data is input to the vibration characteristic learning module 20; (d) comparing the stiffness output in the step (c1) with the magnitude of a danger level already set; (e1) comparing the stiffness output in the step (d1) with the magnitude of the danger level already set; (e2) if the output stiffness is less than the previously set danger level, comparing a cumulative impact amount applied to the building during a predetermined unit period with a resilience threshold of the building in the step (d1); (f1) if the cumulative impact amount is equal to or less than the resilience threshold of the building in the step (e2), repeating the steps (c1) to (e2); and (f2) if the cumulative impact amount exceeds the resilience threshold of the building in the step (e2), an alarm providing module 40 provides an alarm to a building user, an adjacent building management unit, and an infrastructure management unit.

[0021] In addition, preferably, the learning information generation module 10 according to an embodiment of the present disclosure further includes a stiffness measurement unit 110, a pre-processing unit 130, an abnormal climate and disaster information input unit 140, and a utilization information measurement unit 150, and the step (a) includes: (a1) a step in which the stiffness measurement unit 110 collects building stiffness at a normal point A and a deteriorated point B; (a2) a step in which the vibration sensor 120 collects vibration data at the normal point A and the deteriorated point B; (a31) a step in which the pre-processing unit 130 performs FFT (fast fourier transformation) on the vibration data to generate kurtosis; and (a32) a step in which the pre-processing unit 130 uses the vibration data to generate acceleration, RMS, peak, and CF (crest rate) according to time. (a3) a step of calculating a congestion factor (a4) by the abnormal weather and disaster information input unit 140; and (a5) a step of collecting congestion levels by time period and logistics information by time period by the usage information measurement unit 150.

[0022] In addition, preferably, the vibration characteristic learning module 20 according to one embodiment of the present disclosure includes a learning period determination unit 24, and the step (b) includes: (b1) a step in which a prediction model 22 configured to output output information including building stiffness when input information including the kurtosis, acceleration, RMS, peak, CF, congestion level by time period, logistics information by time period, weather information and disaster information is input, is learned by the vibration characteristic learning module 20; and (b2) a step in which the learning period determination unit 24 calculates an information loss value (reconstruction loss) using the input information and the output information, and adjusts the learning period using the information loss value.

[0023] Also, preferably, the method according to one embodiment of the present disclosure further includes, after step (b2), a step (b3) in which the correction module 30 calculates a relational equation between the stiffness of the building output in step (b1) and the building collected in step (a1); and, after step (c1), a step (c2) in which the correction module 30 corrects the stiffness of the building output in step (c1) using the relational equation calculated in step (b3).

[0024] Also, preferably, the stiffness measurement unit 110 according to one embodiment of the present disclosure includes a first stiffness measurement device 112 configured to measure a cavity formed inside a wall of a building, the depth of the cavity and the stiffness of the building; and a second stiffness measurement device 114 configured to measure the stiffness of the building.

[0025] Also, preferably, the first stiffness measuring device 112 according to an embodiment of the present disclosure is an ultrasonic flaw detector, and the second stiffness measuring device 114 is a Schmidt hammer.

[0026] In addition, preferably, according to one embodiment of the present disclosure, when it is determined that an outlier occurs in the raw data of acceleration, RMS, peak, and CF calculated in step (a32), data greater than n·σ (where n is a positive real number) is excluded from the input information.

[0027] Also, preferably, the input information according to an embodiment of the present disclosure further includes a type of building.

[0028] In addition, preferably, according to one embodiment of the present disclosure, the cumulative impact amount is calculated by an impact amount calculation unit, and the impact amount calculation unit calculates the cumulative impact amount using information collected by the abnormal weather and disaster information input unit 140 and the usage information measurement unit 150.

[0029] Moreover, according to one embodiment of the present disclosure, there is provided an apparatus for preventing serious disasters associated with the collapse of a building by predicting the stiffness of the building using information including vibration data collected by a vibration sensor 120 attached to the building, the apparatus including: a learning information generation module 10 configured to collect input information including vibration data collected by the vibration sensor 120 and output information including the stiffness of the building collected by a stiffness measurement unit 110; a vibration characteristic learning module 20 configured to learn a predictive model 22 configured to output the output information when the input information is input; a correction module 30 configured to correct the stiffness of the building output by the vibration characteristic learning module 20; and an alarm providing module 40 configured to decide whether to provide an alarm using the stiffness of the building corrected by the correction module 30, and provide the alarm.

[0030] In addition, preferably, the learning information generation module 10 according to one embodiment of the present disclosure includes a pre-processing unit 130 for pre-processing the vibration data, the pre-processing unit 130 including an FFT transformation unit 132 configured to transform the vibration data into a frequency domain to generate kurtosis; and a vibration characteristic calculation unit 134 for calculating acceleration, RMS, peak, and CF (crest factor) according to time from the vibration data; an abnormal weather and disaster information input unit 140 to which meteorological information and disaster information are input; a usage information measurement unit 150 configured to collect building usage information, the usage information measurement unit 150 including a user congestion degree measurement unit 152 configured to measure congestion degree by time period; and a logistics movement information input unit 154 configured to input logistics information by time period; and a learning information database 160 configured to receive and store data from the stiffness measurement unit 110, the vibration sensor 120, the pre-processing unit 130, the meteorological information and disaster information input unit 140, and the usage information measurement unit 150.

[0031] Also, preferably, the input information according to an embodiment of the present disclosure includes the acceleration, RMS, peak, CF, kurtosis, weather information, disaster information, congestion degree by time period, and logistics information by time period.

[0032] Also, preferably, the vibration characteristic learning module 20 according to one embodiment of the present disclosure includes a learning period determination unit 24, which adjusts the learning period using an information loss value (reconstruction loss) between the input information and the output information.

[0033] Also, preferably, the stiffness measurement unit 110 according to one embodiment of the present disclosure includes a first stiffness measurement device 112 configured to measure a cavity formed inside a wall of a building, the depth of the cavity and the stiffness of the building; and a second stiffness measurement device 114 configured to measure the stiffness of the building. Effect of the Invention

[0034] As described above, according to this embodiment, the stiffness of a building can be predicted using information collected by the vibration sensor, thereby making it possible to detect abnormal behavior of the building. This has the effect of enabling measures to be taken before the destruction or collapse of the building occurs.

[0035] In addition, since the usage information measurement unit and the abnormal weather and disaster information input unit are configured to collect information using already input materials or in cooperation with an external database, the effect is that the user actually only needs to input the vibration data collected by the vibration sensor.

[0036] In addition, by further using the Schmidt hammer in an environment where the reliability of the results of the ultrasonic flaw detector is low, there is an effect of improving the reliability of the results.

[0037] In addition, by selecting the Schmidt hammer as an alternative in environments where it is difficult to use ultrasonic flaw detectors, it is possible to measure the internal strength of concrete in any environment.

[0038] In addition, when outliers are detected in the acceleration, RMS, peak, and CF of the input data, values ​​of 1σ or more are excluded. This has the effect of eliminating data that may be treated as noise from the obtained information, thereby obtaining high-quality training data.

[0039] In addition, the present disclosure has the advantage of being able to predict the rigidity of a building with greater accuracy by taking into account climatic changes and disasters that affect concrete.

[0040] In addition, since the users who affect the concrete are also taken into account, the rigidity of the building can be predicted with greater accuracy.

[0041] In addition, the method takes into account the load on the structure due to the load and flow of the loading logistics and transportation system that affect concrete, which has the effect of making it possible to predict the stiffness of a building with greater accuracy.

[0042] In addition, there is an effect that the learning accuracy is further improved by labeling the vibration data.

[0043] In addition, it is possible to prevent false alarms by considering the resilience of a building's rigidity, as it is determined that a temporary weakening of the building's rigidity will not actually cause the building to be destroyed or collapsed, which has the effect of preventing unnecessary evacuations and the creation of anxiety.

[0044] In addition, the alarm provided to users of the building allows them to escape from the building before the building collapses, thereby minimizing human damage.

[0045] In addition, an alarm can be sent to users and personnel stationed in neighboring buildings around the building being measured, which helps to minimize secondary damage.

[0046] In addition, secondary damage can be minimized by providing alarms to agencies and ministries such as telecommunications companies, Korea Electric Power Corporation, fire departments, Korea Electric Power Corporation, and Korea City Gas Corporation. [Brief description of the drawings]

[0047] [Figure 1] FIG. 1 is a block diagram of a serious disaster prevention system according to one embodiment of the present disclosure. [Diagram 2] 1 is a flowchart of a method for preventing a serious disaster according to an embodiment of the present disclosure. [Diagram 3] FIG. 2 is a block diagram of a learning information generation module according to one embodiment of the present disclosure. [Figure 4] 1 illustrates a data flow according to one embodiment of the present disclosure. [Diagram 5] 1 is a diagram illustrating data processed by a vibration sensor and a preprocessing unit according to one embodiment of the present disclosure. [Figure 6] 1 is a diagram illustrating data processed by a usage information measurement unit according to an embodiment of the present disclosure; [Figure 7] This is to explain Kurtosis according to one embodiment of the present disclosure. [Figure 8] 1 is a comparison between the values ​​predicted by a major disaster prevention system according to an embodiment of the present disclosure and the actual values. [Figure 9] 1 is for illustrating the extent to which an alarm delivery module according to one embodiment of the present disclosure provides an alarm. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0048] Some embodiments of the present disclosure will be described in detail below with reference to the drawings. When referring to components in each drawing, it should be noted that the same components are referred to by the same reference numerals as much as possible when they are displayed in other drawings. In addition, when describing the present disclosure, if a specific description of a related publicly known configuration or function is deemed to obscure the gist of the present disclosure, the detailed description will be omitted.

[0049] In describing components of the embodiments of the present disclosure, reference numerals such as 1, 2, i), ii), a), b) may be used. Such reference numerals are merely for distinguishing the components from other components, and do not limit the essence, sequence, order, etc. of the components. In this specification, when a part "includes" or "has" a certain component, it does not mean that the part may further include other components, but does not mean that the part excludes other components, unless otherwise specified.

[0050] In this disclosure, "rigidity of a building" refers to the internal strength in the vicinity of where the vibration sensor is attached. Since buildings are generally constructed of concrete, it is preferable to understand the rigidity of a building in this disclosure as the concrete strength at the point where the vibration sensor is attached. In this disclosure, the explanation is based on the assumption that the unit of rigidity of a building is Mpa.

[0051] In addition, in this disclosure, "resilience of a building" means the property of a building recovering to its original rigidity after a concentrated force is applied to a local part of the building for a short period of time and the rigidity of the building temporarily decreases. Hereinafter, the cumulative impact amount that exceeds the resilience and does not recover to its original rigidity any more is referred to as the "resilience threshold of a building."

[0052] In addition, in this disclosure, the serious disaster prevention system 1 can be applied to concrete facilities other than buildings, including retaining walls, bridges, and dams, but for ease of explanation, the following description will be given assuming that the system is applied to buildings.

[0053] Overview of Major Disaster Prevention System FIG 1 is a block diagram of a system for preventing a serious disaster according to an embodiment of the present disclosure. FIG 2 is a flowchart of a method for preventing a serious disaster according to an embodiment of the present disclosure.

[0054] According to an embodiment of the present disclosure, when vibration data collected by one or more vibration sensors attached to a building is input to the serious disaster prevention system 1, the serious disaster prevention system 1 is configured to output the stiffness of the building, more specifically, the internal strength (preferably in MPa) of the concrete constituting the building. In addition, when the predicted stiffness of the building is determined to be equal to or higher than a danger level and a serious disaster is predicted, an alarm is provided, thereby minimizing the number of victims from the serious disaster. That is, according to the present disclosure, abnormal behavior of the building can be detected by predicting the stiffness of the building using information collected by the vibration sensor. This has the effect of enabling measures to be taken before the destruction or collapse of the building occurs.

[0055] To this end, a serious disaster prevention system 1 according to an embodiment of the present disclosure includes all or some of a learning information generating module 10, a vibration characteristic learning module 20, a correction module 30, and an alarm providing module 40.

[0056] The learning information generating module 10 is configured to generate information necessary for the vibration characteristic learning module 20 to learn. For this purpose, the learning information generating module 10 includes all or a part of a stiffness measuring unit 110, a vibration sensor 120, a preprocessing unit 130, an abnormal climate and disaster information input unit 140, a usage information measuring unit 150, and a learning information database 160. A detailed description of the generation of learning information by the learning information generating module 10 will be given in detail with reference to Figs. 3 to 7.

[0057] The vibration characteristic learning module 20 is configured to learn a prediction model capable of predicting the stiffness of a building, using the information generated and collected by the learning information generation module 10. The configuration and functions of the vibration characteristic learning module 20 will be described in detail with reference to FIG.

[0058] The correction module 30 is configured to approximately correct the output stiffness of the building to the value of the stiffness of the actual building, using the correlation between the stiffness of the building output by the vibration characteristic learning module 20 and the stiffness of the actual building. The function of the correction module 30 will be described in detail with reference to FIG. 8.

[0059] The alarm providing module 40 is configured to provide an alarm to building users, management offices of adjacent buildings, and infrastructure facilities related to the building being measured when it is determined that the corrected building stiffness has reached a predetermined danger level (threshold). The alarm providing module 40 will be described in detail with reference to FIG. 9.

[0060] With reference to FIG. 2, the overall sequence of a method for predicting a severe disaster using a severe disaster prediction system according to an embodiment of the present disclosure will be described.

[0061] Prediction model input information is collected by the vibration sensor 120, the usage information measurement unit 150, and the abnormal weather and disaster information input unit 140, and prediction model output information is collected by the stiffness measurement unit 110 (S200).

[0062] The vibration characteristic learning module 20 learns the prediction model 22 (see FIG. 4) using all or part of the information collected in step S220 (S210). At this time, the prediction model 22 is configured to output the stiffness of the building when information including information sensed by the vibration sensor 120 is input.

[0063] When the learning is completed for a predetermined learning period, new sensing information that was not used in the learning of the prediction model is input to the prediction model 22 by the vibration sensor 120, and the stiffness of the building (specifically, the internal strength of concrete) is output (S220). Meanwhile, the usage information measurement unit 150 and the abnormal weather and disaster information input unit 140 are configured to use already input data or to collect information in cooperation with an external database (e.g., a database of the Japan Meteorological Agency), so that the user actually only needs to input the vibration data collected by the vibration sensor. At this time, it is preferable that the output stiffness of the building can be corrected by the correction module 30, but it should be noted that this can be appropriately selected according to the needs of the designer.

[0064] Then, the output and / or corrected building stiffness is compared with a pre-set danger level (threshold). If it is determined in step S230 that the output building stiffness is equal to or greater than the pre-set danger level, the process returns to step S220. That is, this is to prevent an alarm from sounding because the building stiffness is greater than the danger level.

[0065] If it is determined that the outputted building stiffness is lower than the previously set danger level (threshold), the alarm providing module 40 compares the accumulated impact accumulated during the unit period with the building resilience threshold (S240). If it is determined in step S240 that the accumulated impact is equal to or lower than the building resilience threshold, the process returns to step S220.

[0066] If it is determined in step S240 that the cumulative impact is greater than the resilience of the building, the alarm providing module 40 provides an alarm (S250). In this case, the alarm providing module 40 can provide an alarm to building users, management offices of adjacent buildings, and infrastructure facilities related to the building being measured.

[0067] Steps S200 to S220 will be described in detail below.

[0068] Data collection and preprocessing for training Fig. 3 is a block diagram of a learning information generating module according to an embodiment of the present disclosure. Fig. 5 is for explaining data processed by a vibration sensor and a pre-processing unit according to an embodiment of the present disclosure. Fig. 6 is for explaining data processed by a usage information measuring unit according to an embodiment of the present disclosure. Fig. 7 is for explaining Kurtosis according to an embodiment of the present disclosure.

[0069] The types of data for learning will be described with reference to FIG. 3 and FIG. 5 to FIG.

[0070] Referring to FIG. 3, the learning information generation module 10 according to one embodiment of the present disclosure includes all or part of a stiffness measurement unit 110, a vibration sensor 120, a pre-processing unit 130, an abnormal weather and disaster information input unit 140, a usage information measurement unit 150, and a learning information database 160.

[0071] The stiffness measurement unit 110 is installed or attached to a portion of a building and is configured to measure the stiffness of the building.

[0072] The stiffness measuring unit 110 can be attached to a normal point A and a deteriorated point B, respectively. Here, the normal point A and the deteriorated point B are points determined by the result of a precise safety diagnosis. A precise safety diagnosis means a building safety diagnosis performed according to standards provided by laws and regulations. The result of the precise safety diagnosis is stored in an external database (not shown) that can be accessed by the building owner or stakeholders of the building, and the stiffness measuring unit 110 is attached to the normal point A that is determined to be normal and the deteriorated point B that is determined to have deteriorated structural strength due to aging using the result of the precise safety diagnosis.

[0073] The stiffness measuring section 110 includes an ultrasonic flaw detector 112 and a Schmidt hammer 114 .

[0074] The ultrasonic flaw detector 112 is configured to transmit ultrasonic waves to the building to which it is attached and to receive the reflected ultrasonic waves. In addition, the ultrasonic flaw detector 112 can detect cavities formed inside the wall of the building by comparing the transmitted ultrasonic waves with the reflected ultrasonic waves. In addition, the ultrasonic flaw detector 112 can measure the depth of the cavities formed inside the wall of the building and the rigidity of the building.

[0075] The Schmidt Hammer 114 is configured to test the internal strength of concrete by impacting the wall to which it is attached and measuring the rebound to estimate the internal strength.

[0076] Meanwhile, the stiffness measuring unit 110 according to the present disclosure basically includes an ultrasonic flaw detector 112, and may selectively or additionally include the ultrasonic flaw detector 112 and the Schmidt hammer 114. This is because by selecting the Schmidt hammer 114 as an alternative in an environment where it is difficult to use the ultrasonic flaw detector 112, it is possible to measure the internal strength of concrete in any environment. Although the ultrasonic flaw detector 112 generally has higher accuracy than the Schmidt hammer 114, the Schmidt hammer 114 has the advantage that it can be applied to various places because its operation method is simple and inexpensive. Therefore, in some limited environments, the Schmidt hammer 114 can be used instead of the ultrasonic flaw detector 112.

[0077] The vibration sensors 120 are configured to collect vibration data and, for this purpose, are attached at the normal point A and the degraded point B, respectively.

[0078] The pre-processing unit 130 is configured to process the vibration data collected by the vibration sensor 120. To this end, the pre-processing unit 130 includes an FFT transformation unit 132 and a vibration characteristic calculation unit 134.

[0079] The vibration data is information about acceleration according to time (see FIG. 5). The FFT transform unit 132 is configured to transform the vibration data from the time domain to the frequency domain to generate frequency data. The frequency data processed by the FFT transform unit 132 may be acceleration information with respect to frequency or amplitude information with respect to frequency.

[0080] The acceleration information with respect to frequency may be expressed in the form of a graph shown below the FFT transform unit 132 in Fig. 5. The amplitude information with respect to frequency may be expressed in the form of a kurtosis graph shown in Fig. 7.

[0081] The vibration characteristic calculation unit 134 is configured to calculate acceleration, RMS (root means square), peak, and CF (crest factor) using the vibration data. The acceleration, RMS, peak, and CF can be understood using the graphs shown below the vibration characteristic calculation unit 134 in FIG.

[0082] When the vibration characteristic calculation unit 134 determines that data that is judged to be close to noise in the collected acceleration, RMS, peak, and CF, i.e., abnormal values ​​exist in the data, it is possible to construct good quality data by excluding the data through a specific range. The acceleration, RMS, peak, and CF are information confirmed from the vibration data collected during the second period in units of a first period, and it is preferable to exclude values ​​that fall outside the specific range. Here, the first period may be, for example, 10 seconds, the second period may be 10 minutes, and the specific range may be 1σ. However, it should be noted that these are merely exemplary values, and the length of the first period and the second period and the specific range can be appropriately changed by the user. In this way, values ​​of the acceleration, RMS, peak, and CF that are equal to or greater than a specific range (for example, 1σ) are excluded, and thus, there is an effect that data that may be treated as noise, i.e., abnormal values, are excluded from the obtained information, and good quality learning data is obtained.

[0083] The abnormal climate and disaster information input unit 140 is configured to input weather changes and disasters that affect concrete deterioration. When events such as extreme cold, extreme heat, heavy rain, earthquakes, and typhoons occur that could not be considered at the time of design, structures made of concrete will be affected. Weather change and disaster information is input to the abnormal climate and disaster information input unit 140 on a set period basis, for example, on a weekly or quarterly basis. The present disclosure has the effect of being able to predict the stiffness of a building with higher accuracy because it takes into account weather changes and disasters that affect concrete.

[0084] The usage information measuring unit 150 is configured to receive usage information of the building according to time. To this end, the usage information measuring unit 150 includes a user congestion degree measuring unit 152 and a logistics movement information input unit 154.

[0085] The user congestion level measurement unit 152 is configured to measure the congestion level by time period (see (a) of FIG. 6). (a) of FIG. 6 shows the average number of users at Incheon Airport by time period from July to August 2022. Referring to (a) of FIG. 6, it can be seen that the largest number of users is between 8:00 and 20:00. Impacts according to user loads also affect the rigidity of buildings. The present disclosure has the effect of being able to predict the rigidity of buildings with higher accuracy since it also takes into account users who affect concrete.

[0086] The logistics movement information input unit 154 is configured to input logistics information by time period (see (b) of FIG. 6). (b) of FIG. 6 shows the weight average of cargo placed at Incheon Airport by time period from July to August 2022. Referring to (b) of FIG. 6, it can be seen that the weight of cargo is the heaviest between 9:00 and 20:00. In other words, for vibration data observed between 9:00 and 20:00, if it is an office or residential building, the weight of cargo entering and leaving the building is not important, but if it is a logistics warehouse or airport, the weight of cargo entering and leaving the building affects the rigidity of the building. The present disclosure has the effect of predicting the rigidity of a building with higher accuracy because it also takes into account the weight of logistics that affects concrete.

[0087] Using the congestion degree by time period, the early morning time period (Ta) with the least congestion and the daytime time period (Tb) with the highest congestion can be extracted. According to the example of FIG. 6, Ta may be from 5:00 to 6:00, and Tb may be from 17:00 to 18:00. Taking into consideration the time period, the vibration data for the time period can be labeled and input to the prediction model 22. Similarly, the vibration data may be further labeled using logistics information by time period. Labeling the vibration data has the effect of further improving the learning accuracy.

[0088] Meanwhile, the information shown in FIG. 6 is merely an example for explaining the degree of user congestion and the logistics movement information, and the information obtained may differ depending on the type of building, etc.

[0089] The learning information database 160 is configured to receive and store the data collected by the stiffness measurement unit 110, the vibration sensor 120, the preprocessing unit 130, the abnormal weather and disaster information input unit 140, and the usage information measurement unit 150. In this case, the learning information database 160 preferably stores the received information in a data set. For example, the learning information database 160 can vectorize the input information and store it.

[0090] The type of building may be further input to the learning information database 160. Examples of the types of buildings include general buildings, department stores, supermarkets, public facilities (e.g., sports facilities, libraries, etc.), airports, schools, concrete retaining walls, and infrastructure facilities in industrial complexes.

[0091] The learning information database 160 can transmit the stored data set to the vibration characteristic learning module 20 upon request of the vibration characteristic learning module 20 .

[0092] Learning model input data, output data, and prediction results FIG. 4 illustrates the data flow according to one embodiment of the present disclosure.

[0093] 4, the vibration characteristic learning module 20 is configured to learn a prediction model 22. The prediction model 22 receives as input data the acceleration, RMS, peak, CF, and kurtosis collected or processed by the vibration sensor 120 and the preprocessing unit 130, the congestion degree by time period measured by the user congestion degree measuring unit 152, the logistics information by time period input to the logistics movement information input unit 154, and the weather information and disaster information input to the abnormal weather and disaster information input unit 140, and receives as output data the rigidity of the building measured by the rigidity measuring unit 110.

[0094] Preferably, the prediction model 22 can be realized by an unsupervised learning or reinforcement learning model. However, it is not necessarily limited to this, and may be realized by any of an artificial neural network (ANN), a convolutional neural network (CNN), a recurrent neural networks (RNN), a Bayesian network, a Long Short-Term Memory (LSTM), a generative adversarial network (GAN), a Transformer, and a decision tree, and is not necessarily limited to the above models, and may be realized by any type of prediction model.

[0095] Anomaly detection in time-series vibration data involves detecting anomalies in observed value points, anomalies due to changes in data trends, etc., and measuring the anomaly score, which indicates the degree of anomaly, for each time-series interval. The methods for measuring the anomaly score for each method of the artificial intelligence prediction models mentioned above can be divided into the categories of "Reconstruction error," "Prediction error," and "Dissimilarity." Here, we measure the anomaly score that is most suitable for each prediction model.

[0096] For example, when using LSTM and Autoencoder, the prediction model operates by encoding time series data into a low-dimensional space and then measuring the reconstruction error by decoding the encoded vector again. The reconstruction error thus measured is assumed to be small for normal data and large for abnormal data, and abnormal data is detected.

[0097] The vibration characteristic learning module 20 may arbitrarily determine the learning period, but preferably, may adjust the learning period in order to obtain a more accurate result. For this purpose, the vibration characteristic learning module 20 may further include a learning period determination unit 24.

[0098] The learning period determination unit 24 can reset the learning period in consideration of the stiffness of the building output by the prediction model 22 and the reconstruction loss for the difference between the input data and the output data input to the prediction model 22. Specifically, the learning period can be adjusted until the degree of anomaly in the information loss value falls within a preset threshold range. In this way, an appropriate learning period can be set regardless of the type of learning model that the prediction model 22 is realized by, and compatibility with various models of already given data can be increased.

[0099] FIG. 8 compares the values ​​predicted by the major disaster prevention system according to one embodiment of the present disclosure with the actual values.

[0100] Referring to FIG. 8, it can be seen that there is a high correlation between the actual stiffness of the building (see the actual axis) and the stiffness of the building output by the vibration characteristic learning module 20 (see the predicted axis).

[0101] However, since there may be a certain degree of difference between the stiffness of the actual building and the stiffness of the output building, it is necessary to correct the stiffness of the building output by the prediction model 22. For this purpose, the major disaster prevention system 1 of the present disclosure can correct the stiffness of the output building via the correction module 30 so that the stiffness of the output building approximates the stiffness of the actual building. For example, the correction module 30 can correct the stiffness of the output building using linear regression. However, it should be noted that this is merely an example, and a more appropriate correction formula can be selected by the user.

[0102] Hereinafter, a method for determining whether to provide an alarm using the outputted stiffness of the building will be described in detail (see steps S230 to S250 in FIG. 2).

[0103] Alarm delivery method According to step S230 of FIG. 2, the alarm providing module 40 may determine whether the outputted building rigidity is greater than a pre-set danger level (threshold). In this case, the pre-set danger level may be a structural safety standard strength designated by regulations regarding building structural standards, etc. However, it is not necessarily limited thereto, and may be adjusted according to the management level of a safety manager or a building owner.

[0104] If it is determined that the outputted building rigidity is equal to or higher than the preset danger level, the alarm providing module 40 decides not to provide an alarm, and returns to the step of collecting vibration data using the vibration sensor 120.

[0105] According to step S240 of FIG. 2, if it is determined that the outputted building stiffness is less than the previously set danger level, the alarm providing module 40 determines whether the accumulated impact is greater than the building resilience threshold.

[0106] After a concentrated force is applied to a local part of a building for a short period of time, the rigidity of the building may temporarily decrease. If such impacts are accumulated, the building will lose its resilience. Here, the magnitude of the impact at which the building loses its resilience is called the resilience threshold. The magnitude of the accumulated impact is the accumulation of the magnitude of the impacts applied to the building during a preset period, i.e., a unit period. For this purpose, a separately provided impact amount calculation unit (not shown) can calculate the accumulated impact amount using information measured or input by the abnormal weather and disaster information input unit 140 and the usage information measurement unit 150.

[0107] For example, when the abnormal weather and disaster information input unit 140 confirms that an earthquake has occurred near a building to be measured, the impact amount calculation unit can calculate the magnitude of the impact applied to the building using the earthquake's seismic intensity, the distance to the epicenter, and the duration of the earthquake. Alternatively, when the usage information measurement unit 150 detects more than twice the number of people and goods available in the building in one hour, the impact amount calculation unit can calculate the magnitude of the impact applied to the building by the weight of the users and goods.

[0108] If it is determined in step S240 that the magnitude of the cumulative impact is equal to or less than the resilience threshold of the building, the alarm providing module 40 decides not to provide an alarm. Therefore, the process returns to the step of collecting vibration data using the vibration sensor 120. In other words, it is determined that the result of temporary weakening of the building's rigidity will not actually cause the destruction or collapse of the building, and a false alarm can be prevented. This has the effect of preventing unnecessary evacuations and the creation of anxiety.

[0109] If it is determined in step S240 that the magnitude of the cumulative impact amount is greater than the resilience threshold of the building, the alarm providing module 40 determines to provide an alarm in step S250.

[0110] Here, the alarms provided by the alarm providing module 40 include alarms provided to users in the building, alarms provided to adjacent building management departments, and alarms provided to infrastructure facilities related to the building.

[0111] The alarm provided to users in the building can be provided, for example, through a speaker or a display. Alternatively, a push alarm can be provided through an application (for example, a department store application) installed on the terminal of the user in the building. This allows users using the building to escape from the building before the building collapses, thereby minimizing human damage.

[0112] The alarm provided to the adjacent building management unit may be provided via a terminal or the like provided in a management unit provided in the adjacent building. Here, the adjacent building may include buildings and facilities facing the measurement target building as shown in FIG. 9. If the measurement target building collapses, secondary damage may occur in the adjacent building due to falling rocks and dust generated during the collapse. Meanwhile, when an alarm is provided to the adjacent building management unit, it is possible to prepare for damage caused by falling rocks and impacts, and an alarm can be provided to users and resident personnel of the adjacent building, thereby minimizing secondary damage.

[0113] Alarms provided to infrastructure facilities related to buildings can be provided to, for example, overhead high-voltage lines, communication facilities, high-pressure tanks, substations, and agencies, ministries, and facilities in charge of city gas and water supply. Damage to infrastructure such as electricity, communication, gas, and water supply can cause widespread damage. Therefore, by providing alarms to agencies and ministries such as communication companies, Korea Electric Power Corporation, fire departments, Korea Electric Power Corporation, and Korea Gas Corporation, secondary damage can be minimized.

[0114] The above description is merely an illustrative example of the technical idea of ​​the present embodiment, and various modifications and variations are possible within the scope of the essential characteristics of the present embodiment, if one has ordinary knowledge in the technical field to which the present embodiment belongs. Therefore, the present embodiment is intended to explain, not to limit, the technical idea of ​​the present embodiment, and the scope of the technical idea of ​​the present embodiment is not limited by such an embodiment. The scope of protection of the present embodiment should be interpreted according to the claims below, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the rights of the present embodiment. [Explanation of symbols]

[0115] 1. Major disaster prevention system 10. Learning information generation module 20. Vibration Characteristics Learning Module 30 Correction module 40 Alarm provision module 110...Rigidity measurement section 120 ···Vibration sensor 130 Pre-processing section 140: Extreme weather and disaster information input section 150...Usage information measurement section 160 ···Learning Information Database

Claims

1. A method for preventing a serious disaster associated with a collapse of resilience of a building by predicting the stiffness of the building using information including vibration data collected by a vibration sensor (120) attached to the building, comprising: (a) a learning information generation module (10) collects input information including the vibration data and output information including the stiffness of the building; (b) a vibration characteristic learning module (20) learning a predictive model (22) using the input information and the output information; (c1) when the vibration data is input to the vibration characteristic learning module (20), the stiffness of the building is output; (d) comparing the stiffness output in the step (c1) with a previously set danger level; (e1) repeating steps (c) to (d1) if the output stiffness is equal to or greater than the preset danger level in step (d1); (e2) in the step (d1), if the output stiffness is less than the previously set danger level, a step of comparing a cumulative impact amount applied to the building during a predetermined unit period with a resilience threshold value of the building; (f1) if the cumulative impact amount is equal to or less than the resilience threshold of the building in step (e2), steps (c1) to (e2) are repeated; and (f2) in the step (e2), when the cumulative impact amount exceeds the resilience threshold of the building, an alarm providing module (40) provides an alarm to a building user, an adjacent building management unit, and an infrastructure management unit; A method comprising:

2. The learning information generation module (10) includes a stiffness measurement unit (110), a pre-processing unit (130), an abnormal weather and disaster information input unit (140), and a usage information measurement unit (150), The step (a) comprises: (a1) a step in which the stiffness of the building at normal points (A) and deteriorated points (B) is collected by the stiffness measurement unit (110); (a2) collecting the vibration data of the normal point (A) and the deteriorated point (B) by the vibration sensor (120); (a31) the pre-processing unit (130) subjects the vibration data to FFT (fast Fourier transformation) processing to generate kurtosis; (a32) the pre-processing unit (130) uses the vibration data to calculate acceleration, RMS, peak and crest factor (CF) as a function of time; (a4) a step of collecting meteorological information and disaster information by the abnormal weather and disaster information input unit (140); and (a5) collecting congestion levels and logistics information by time period by the usage information measuring unit (150); The method of claim 1 , comprising:

3. The vibration characteristic learning module (20) includes a learning period determination unit (24), The step (b) comprises: (b1) a step in which a prediction model (22) configured to output the output information including the stiffness of the building when the input information including the kurtosis, the acceleration, the RMS, the peak, the CF, the congestion degree by time period, the logistics information by time period, the weather information, and the disaster information is input, is learned by the vibration characteristic learning module (20); and (b2) the learning period determination unit (24) calculates an information loss value (reconstruction loss) using the input information and the output information, and adjusts a learning period using the information loss value; The method of claim 2 , comprising:

4. After the step (b2), (b3) a correction module (30) calculates the stiffness of the building outputted in the step (b1) and the relational equation of the building collected in the step (a1); and After the step (c1), (c2) the correction module (30) corrects the stiffness of the building outputted in the step (c1) using the relational expression calculated in the step (b3); The method of claim 3 further comprising:

5. The stiffness measurement unit (110) includes a first stiffness measurement device (112) configured to measure a cavity formed inside a wall of the building, the depth of the cavity, and the stiffness of the building; and a second stiffness measurement device (114) configured to measure the stiffness of the building; The method of claim 2 , comprising:

6. The first stiffness measuring device (112) is an ultrasonic flaw detector, The method of claim 5, wherein the second stiffness measuring device (114) is a Schmidt hammer.

7. 3. The method of claim 2, wherein, when an abnormal value is determined to occur in the raw data of the acceleration, the RMS, the peak, and the CF calculated in step (a32), data of n·σ or more (where n is a positive real number) is excluded from the input information.

8. The method of claim 2 , wherein the input information further comprises a type of the building.

9. The cumulative impact amount is calculated by an impact amount calculation unit, The method according to claim 2 , wherein the impact amount calculation unit calculates the cumulative impact amount using information collected by the abnormal weather and disaster information input unit (140) and the usage information measurement unit (150).

10. A device for preventing a serious disaster caused by a collapse of the resilience of a building by predicting the stiffness of the building using information including vibration data collected by a vibration sensor (120) attached to the building, comprising: a learning information generation module (10) configured to collect input information including the vibration data collected by the vibration sensor (120) and output information including the stiffness of the building collected by a stiffness measurement unit (110); a vibration characteristic learning module (20) configured to learn a predictive model (22) configured to output the output information when the input information is received; A correction module (30) configured to correct the stiffness of the building output by the vibration characteristic learning module (20); and an alarm providing module (40) configured to determine whether to provide an alarm using the stiffness of the building corrected by the correction module (30) and to provide the alarm; 13. An apparatus comprising:

11. The learning information generation module (10) A pre-processing unit (130) for pre-processing the vibration data, an FFT transform unit (132) configured to transform the vibration data into the frequency domain to generate a kurtosis; and a pre-processing unit (130) including a vibration characteristic calculation unit (134) for calculating acceleration, RMS, peak, and CF (crest factor) according to time from the vibration data; an extreme weather and disaster information input unit (140) into which meteorological and disaster information is input; and A usage information measurement unit (150) configured to collect usage information of the building, A user congestion level measurement unit (152) configured to measure congestion levels by time period; and A utilization information measurement unit (150) including a logistics movement information input unit (154) configured to input logistics information by time period; and a learning information database (160) configured to receive and store data from the stiffness measurement unit (110), the vibration sensor (120), the pre-processing unit (130), the meteorological and disaster information input unit (140), and the utilization information measurement unit (150); The apparatus of claim 10 , comprising:

12. The device according to claim 11 , wherein the input information includes the acceleration, the RMS, the peak, the CF, the kurtosis, the weather information, the disaster information, the congestion degree by time period, and the logistics information by time period.

13. The vibration characteristic learning module (20) includes a learning period determination unit (24), The apparatus of claim 12, wherein the learning period determination unit (24) adjusts the learning period using an information loss value (reconstruction loss) between the input information and the output information.

14. The stiffness measurement unit (110) includes a first stiffness measurement device (112) configured to measure a cavity formed inside a wall of the building, the depth of the cavity, and the stiffness of the building; and a second stiffness measurement device (114) configured to measure the stiffness of the building; The apparatus of claim 10 , comprising:

Images