Method and device for determining the maximum usable distance of strong vibration acceleration recording

Abstract

The present application provides a method and device for determining the maximum available distance of strong vibration acceleration records, and relates to the field of earthquake engineering technology. The method of the present application includes obtaining strong vibration acceleration records of target type instruments; based on the data quality group to which each record belongs determined in a factual standard manner, statistical comparative analysis of the records is performed to determine the ground peak acceleration threshold used for rapid quality classification of data records obtained by monitoring of this type of instrument; using the earthquake motion attenuation model to construct a mapping relationship of ground peak acceleration attenuation with distance, bringing the ground peak acceleration threshold into the mapping relationship, and determining the maximum available distance for obtaining data obtained by monitoring of the target type of instrument. By applying the technical solution of the present application, it is possible to quickly evaluate the quality of data obtained by instruments configured at sites in an actual network, and obtain the maximum available distance of records measured at the site, effectively providing a basic basis for related engineering applications.

Description

Technical Field

[0001] The present application relates to the field of earthquake engineering technology, and in particular to a method and device for determining a maximum available distance for strong vibration acceleration recording. Background Art

[0002] Strong vibration observation is an important basis for studying the characteristics of strong ground motion, earthquake sources and site impacts. In order to effectively carry out strong vibration observation, it is necessary to establish a large-scale, high-density earthquake monitoring and early warning station. For example, the implementation of my country's "National Earthquake Intensity Rapid Reporting and Early Warning Project" has completed the construction of a high-density earthquake early warning network across the country, including more than 5,000 benchmark stations and basic stations, as well as more general stations. In the actual network construction and implementation, compared with strong vibration meters, intensity meters based on micro-electromechanical system accelerometers (MEMS accelerometers) have the advantages of fewer deployment restrictions and low cost. In order to ensure the density of stations, this type of instrument is widely used in general station construction.

[0003] In the process of realizing the present invention, the inventors found that compared with the seismic records obtained by strong motion meters, the MEMS accelerometers in this type of instrument have the characteristics of low precision and high noise level; in practice, a variety of such observation instruments are put into use, and due to the influence of factors such as instrument accuracy and self-noise between different instruments, the available range of observation records of such stations in the network is not the same, which brings many difficulties to the selection of seismic records in engineering seismic resistance and seismic motion modeling, and is not conducive to the development of related earthquake services.

[0004] Therefore, in order to provide a basic basis for related earthquake engineering applications based on strong vibration observation data, how to quickly determine the maximum available distance of strong vibration acceleration records obtained by MEMS acceleration sensors has become a technical problem that needs to be solved urgently.

[0005] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0006] In order to overcome the problems existing in the related art at least to a certain extent, the embodiments of the present application provide a method and device for determining the maximum available distance of strong vibration acceleration records. For the sites determined in the network and the corresponding instrument configurations, a statistical comparison and analysis method is used to determine the ground peak acceleration threshold that can be used for rapid classification of data records, and then determine the maximum available distance of the records measured at the site.

[0007] First aspect

[0008] The present application provides a method for determining a maximum available distance of a strong vibration acceleration record, the method comprising:

[0009] Obtain strong vibration acceleration records of target instruments;

[0010] Based on the data quality group to which each record belongs determined in a factual standard manner, a statistical comparative analysis is performed on the strong vibration acceleration records to determine a ground peak acceleration threshold for rapid quality classification of data records monitored by such instruments;

[0011] A mapping relationship between ground peak acceleration and distance attenuation is constructed using a seismic attenuation model, and the ground peak acceleration threshold is brought into the mapping relationship to determine a maximum available distance for obtaining data monitored by the target type instrument.

[0012] In a possible implementation, the data quality group includes a group of unusable records and at least one group of usable records; the statistical comparative analysis process includes:

[0013] Determining the ground peak acceleration value represented by each of the strong vibration acceleration records;

[0014] For a group pair consisting of a group of unavailable records and a group of available records in the data quality group, within the value range of the ground peak acceleration value, a trial value of the ground peak acceleration threshold for classification of the group pair is selected by traversing with a preset step length, and a binary classification confidence index corresponding to the trial value is determined based on the following expression:

[0015]

[0016] in, represents the binary classification confidence index, A represents the proportion of records correctly classified as the group of available records by the trial value, and R represents the proportion of records correctly classified as the group of unavailable records by the trial value;

[0017] The values ​​of the binary confidence indexes obtained are compared, and the trial value corresponding to the binary confidence index with the largest value is used as the ground peak acceleration threshold for group classification of the available records.

[0018] In a possible implementation, the process of determining the data quality group to which each record belongs in a factual standard manner includes:

[0019] Dividing the time series data represented by the strong vibration acceleration record into signal segments and noise segments;

[0020] Comparing the Fourier amplitude spectra of the signal segment and the noise segment to obtain a signal-to-noise ratio curve representing the signal-to-noise ratio of each frequency point;

[0021] According to the signal-to-noise ratio curve, with the signal-to-noise ratio being greater than or equal to a first threshold as a cutoff condition, determining a low-frequency cutoff frequency and a high-frequency cutoff frequency of the strong vibration acceleration record, and obtaining an available frequency band feature of the strong vibration acceleration record;

[0022] According to the available frequency band characteristics of each strong vibration acceleration record, the data quality group to which the corresponding record belongs is determined.

[0023] In a possible implementation, the first threshold is 3.

[0024] In a possible implementation, within the value range of the ground peak acceleration value, the process of selecting the trial value of the ground peak acceleration threshold for group classification by traversing with a preset step length includes:

[0025] Based on the actual data characteristics recorded in the group pair, intercepting the threshold estimated value range from the value range of the ground peak acceleration value;

[0026] Within the threshold estimation value range, a trial value of the ground peak acceleration threshold for group classification is selected by traversing with a preset step length.

[0027] In a possible implementation, the mapping relationship between the ground peak acceleration and the distance attenuation is constructed by using the earthquake attenuation model, specifically:

[0028] The earthquake motion attenuation model is used to construct the mapping relationship between the attenuation of the ground peak acceleration and the fault distance or the source distance in earthquake events of different magnitudes.

[0029] Second aspect

[0030] The present application provides a device for determining a maximum available distance of a strong vibration acceleration record, the device comprising:

[0031] An acquisition module is used to acquire strong vibration acceleration records of target instruments;

[0032] A first determination module is used to perform statistical comparative analysis on the strong vibration acceleration records based on the data quality group to which each record belongs determined in a factual standard manner, and determine a ground peak acceleration threshold for rapid quality classification of data records monitored by such instruments;

[0033] The second determination module is used to construct a mapping relationship between ground peak acceleration and distance attenuation using a seismic attenuation model, bring the ground peak acceleration threshold into the mapping relationship, and determine the maximum available distance for obtaining data obtained by instrument monitoring of the target type.

[0034] The method for determining the maximum available distance of strong vibration acceleration records provided by the embodiment of the present application includes: obtaining strong vibration acceleration records of target type instruments; based on the data quality group to which each record belongs determined in a factual standard manner, performing statistical comparative analysis on the strong vibration acceleration records, and determining the ground peak acceleration threshold for rapid quality classification of the data records monitored by this type of instrument; using the earthquake motion attenuation model to construct a mapping relationship of ground peak acceleration with distance attenuation, bringing the ground peak acceleration threshold into the mapping relationship, and determining the maximum available distance for obtaining the data monitored by the target type of instrument. By applying the technical solution of the present application, it is possible to perform rapid quality assessment on the data obtained by the instruments configured at the sites in the actual network, and obtain the maximum available distance of the records measured at the sites, which can provide a basic basis for related earthquake engineering applications based on strong vibration observation data, and is conducive to the development of related earthquake service applications.

[0035] Other advantages, objectives, and features of the present application will be described in part in the following description, and in part will be apparent to those skilled in the art based on the following examination and study, or may be taught from the practice of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings are used to provide a further understanding of the technical solution of the present application or the prior art, and constitute a part of the specification. Among them, the accompanying drawings expressing the embodiments of the present application are used together with the embodiments of the present application to explain the technical solution of the present application, but do not constitute a limitation on the technical solution of the present application.

[0037] Figure 1 A flow chart of a method for determining the maximum available distance of strong vibration acceleration recording provided by one embodiment of the present application;

[0038] Figure 2 A schematic diagram illustrating the mapping relationship between the ground peak acceleration and the distance attenuation in one embodiment of the present application;

[0039] Figure 3 A schematic diagram of the structure of a device for determining the maximum available distance for strong vibration acceleration recording provided in one embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other implementation methods obtained by ordinary technicians in the field without making creative work belong to the scope of protection of the present application.

[0041] As described in the background technology, strong vibration observation is an important basis for studying the characteristics of strong ground motion, earthquake sources and site effects. In order to effectively carry out strong vibration observation, it is necessary to establish a large-scale and high-density earthquake monitoring and early warning station. For example, the implementation of my country's "National Earthquake Intensity Rapid Reporting and Early Warning Project" has completed the construction of a high-density earthquake early warning network including more than 5,000 reference stations and basic stations, as well as more general stations across the country. In the actual network construction and implementation, compared with general strong vibration instruments, intensity meters based on micro-electromechanical system acceleration sensors (MEMS acceleration sensors) have the advantages of fewer layout restrictions and low cost. In order to ensure the density of stations, such instruments are widely used in general station construction.

[0042] In the process of realizing the present invention, the inventors found that compared with the seismic records obtained by strong motion meters, the MEMS accelerometers in this type of instrument have the characteristics of low precision and high noise level; in practice, a variety of such observation instruments are put into use, and due to the influence of factors such as instrument accuracy and self-noise between different instruments, the available range of observation records of such stations in the network is not the same, which brings many difficulties to the selection of seismic records in engineering seismic resistance and seismic motion modeling, and is not conducive to the development of related earthquake services.

[0043] Based on this, the present application proposes a method for determining the maximum available distance of strong vibration acceleration records. For the sites determined in the network and the corresponding instrument configurations, a statistical comparison method is used to determine the ground peak acceleration threshold that can be used for rapid classification of data records, and then determine the maximum available distance of the records measured at the site.

[0044] like Figure 1 As shown, in one embodiment, the method for determining the maximum available distance of strong vibration acceleration recording proposed in the present application includes:

[0045] Step S110, obtaining strong vibration acceleration records of the target type.

[0046] For example, in this embodiment, 20,871 strong vibration acceleration records observed by the MEMS intensity meter were obtained.

[0047] Step S120, based on the data quality group to which each record belongs determined in a factual standard manner, statistical comparative analysis is performed on the strong vibration acceleration records to determine the ground peak acceleration threshold for rapid quality classification of the data records obtained by monitoring of this type of instrument.

[0048] It should be noted here that the de facto standard method in step S120 refers to a data quality evaluation method that has been recognized by the public for signal evaluation in the technical field (such as a method of evaluating and classifying based on available frequency band characteristics). In this way, the data obtained in step S110 is evaluated and classified for quality, and the obtained classification result is used as a reference for subsequent comparative analysis. In the present application, the data quality groups here include a group of unusable records and at least one group of available records;

[0049] For example, based on actual needs, in this embodiment, the classification results are divided into four groups. The three groups of available records correspond to records with very wide available frequency bands (or Very Broadband Record, VBBR for short), records with relatively wide available frequency bands (also Broadband Record, BBR for short) and records with limited bandwidth (also Narrowband Record, NBR for short); the group of unavailable records corresponds to records with very narrow available frequency bands that cannot be used for engineering applications (also Rejected Record, REJ for short);

[0050] Specifically, in step S120, the process of statistical comparison and analysis includes:

[0051] Determine the peak ground acceleration represented by each strong vibration acceleration record, referred to as the PGA (Peak Ground Acceleration) value, which is the maximum amplitude of the corresponding record.

[0052] For a group pair consisting of a group with unavailable records and a group with available records in the data quality group, within the value range of the ground peak acceleration value, a trial value PGA of the ground peak acceleration threshold for classification of the group pair is selected by traversing with a preset step length. th-s , and determine the binary confidence index corresponding to the trial value based on the following expression (1):

[0053] (1)

[0054] In expression (1), Represents the binary classification confidence index, A represents the proportion of records correctly classified as the group of available records by the trial value, and R represents the proportion of records correctly classified as the group of unavailable records by the trial value;

[0055] It is easy to understand that, taking the group pair consisting of REJ group and VBBR group as an example, the A in the above process means that PGA>PGA in the VBBR record th-s The cumulative sum of frequencies, R means that PGA<PGA in REJ record th-sThe cumulative sum of frequencies.

[0056] In the above manner, the binary confidence index values ​​under various trial values ​​are obtained, and then the values ​​of the binary confidence indexes obtained are compared. The trial value corresponding to the binary confidence index with the largest value is used as the ground peak acceleration threshold for the group classification of the available records. Taking the group pair consisting of the REJ group and the VBBR group as an example, the ground peak acceleration threshold PGA for VBBR to distinguish REJ is determined. th .

[0057] In this embodiment, the above process is repeated to obtain the thresholds and corresponding confidence indexes of VBBR, BBR, and NBR for distinguishing REJ records. th They are 0.004g, 0.0025g and 0.00125g respectively.

[0058] In relevant practical applications, the records observed by such instruments can be recorded in PGA format. th For quick assessment and classification, such as classifying records with PGA greater than 0.004g as VBBR, records with PGA greater than 0.0025g and less than 0.004g as BBR, records with PGA greater than 0.00125g and less than 0.0025g as NBR, and records with PGA less than 0.00125g as REJ.

[0059] Further, as a specific implementation, in order to improve the trial calculation efficiency, the process of traversing and selecting the trial calculation value of the ground peak acceleration threshold for group classification with a preset step length within the value range of the ground peak acceleration value includes:

[0060] Based on the actual data characteristics recorded in the group alignment, the threshold estimation range is extracted from the range of ground peak acceleration values. Within the threshold estimation range, the trial value of the ground peak acceleration threshold for group classification is selected by traversing with a preset step size. That is, based on actual engineering experience, the actual traversal range is narrowed to improve the trial calculation efficiency.

[0061] After step S120, step S130 is performed to construct a mapping relationship between ground peak acceleration and distance attenuation using a seismic attenuation model, bring the ground peak acceleration threshold into the mapping relationship, and determine the maximum available distance for obtaining data monitored by the target type instrument.

[0062] Specifically, in this step, a mapping relationship between the attenuation of ground peak acceleration and fault distance or source distance in earthquake events of different magnitudes can be constructed using the earthquake attenuation model according to the earthquake attenuation model;

[0063] In practice, based on the needs, the site conditions can be set as general soil sites (i.e., China Class II average sites. According to the research results of Xie et al (2023), the corresponding China Class II average sites are taken as V S30 =359 m / s), and using the earthquake motion attenuation model in this case, determine the mapping relationship under five different magnitudes: M4, M5, M6, M7, and M8;

[0064] For example, in actual implementation, according to the earthquake attenuation model, the 5% quantile PGA (taken as the average value predicted by the earthquake motion model minus the standard deviation within two events) of five different magnitudes (M4, M5, M6, M7, and M8) is plotted as a function of distance, such as Figure 2 The corresponding attenuation curve shown in (the figure shows the curve of attenuation with fault distance, R JB represents the fault distance).

[0065] Then, the ground peak acceleration threshold obtained in step S120 is brought into the corresponding mapping relationship to obtain the corresponding maximum available distance. Figure 2 As shown, the ground motion PGA attenuation curve and the threshold PGA th The result of the intersection is the recommended maximum available distance R max .

[0066] For example, in one embodiment, two sets of threshold PGAs are obtained for two types of MEMS intensity meters with different noise levels used by the China Earthquake Early Warning Network. th , respectively:

[0067] (a) 0.004g, 0.0025g and 0.00125g, applicable to the recording of five types of intensity meters with lower noise levels: Palert Advance, GL-P2B, GZ-GN4, TMA-33 and VH-GL-LDY;

[0068] (b) 0.0079g, 0.005g and 0.0025g, applicable to the recording of two types of intensity meters, MI3000 and REMOS-SIT4, which have higher noise levels;

[0069] The corresponding maximum available distances are shown in Table 1 below:

[0070] Table 1 - Maximum available distances for each magnitude under different thresholds

[0071]

[0072] In Table 1, ① corresponds to the classification results recorded by two intensity meters (MI3000 and REMOS-SIT4); ② corresponds to the classification results recorded by five intensity meters (Palert Advance, GL-P2B, GZ-GN4, TMA-33 and VH-GL-LDY).

[0073] By applying the technical solution of the present application, it is possible to quickly evaluate the quality of the data obtained by the instruments configured at the sites in the actual network, and to obtain the maximum available distance of the records measured at the sites, which can provide a basic basis for related earthquake engineering applications based on strong motion observation data. For example, in practice, combined with the situation shown in Table 1, for site A configured with MI3000, when an M8 earthquake occurs, if site A is more than 325 kilometers away from the earthquake fault, the records generated by its monitoring will become unavailable, which is conducive to the effective implementation of related earthquake services (such as earthquake early warning services).

[0074] To facilitate understanding of the technical solution of the present application, the following briefly describes the implementation method of determining the data quality group to which each record belongs in a factual standard manner in the technical solution of the present application.

[0075] In one embodiment, the process of determining the data quality group to which each record belongs in a de facto standard manner includes:

[0076] The time series data represented by the strong vibration acceleration record is divided into signal segments and noise segments. For example, the P wave arrival result is identified by automatic wave arrival picking, and the time series before the P wave arrival is used as the noise segment, and the position from the beginning of the P wave to the normalized cumulative intensity value equal to 0.9 is used as the signal segment;

[0077] Then, a time-frequency transformation is performed, and the Fourier amplitude spectra of the signal segment and the noise segment are compared to obtain a signal-to-noise ratio curve representing the signal-to-noise ratio of each frequency point;

[0078] Then, according to the signal-to-noise ratio curve, the signal-to-noise ratio SNR is greater than or equal to the first threshold as the interception condition (such as SNR ≥ 3), the low-frequency cutoff frequency and the high-frequency cutoff frequency of the strong vibration acceleration record are determined, and the available frequency band characteristics of the strong vibration acceleration record are obtained;

[0079] According to the available frequency band characteristics of each strong vibration acceleration record, the data quality group to which the corresponding record belongs is determined.

[0080] Further, for example, in one embodiment, fcL is used to represent the low frequency cutoff frequency, and fcH is used to represent the high frequency cutoff frequency, and the classification rules include:

[0081] The records whose available frequency band characteristics satisfied fcL≤0.2Hz and fcH≥20Hz were identified as the data group with a wide available frequency band (i.e., the VBBR group);

[0082] Records that do not meet the VBBR group conditions but meet the available frequency band characteristics of fcL≤0.5Hz and fcH≥10Hz are determined as the data group with a wider available frequency band (i.e., the BBR group);

[0083] The records whose available frequency band characteristics satisfy fcL>0.5Hz or fcH<10Hz, and fcH / fcL>5 are identified as the data group with limited bandwidth (i.e., NBR group);

[0084] The records with signal-to-noise ratio (SNR) < 3 in the range of 0-50 Hz, records with fcH / fcL < 5, and records with unclear seismic phases in the time series were identified as unusable data groups (i.e., REJ group).

[0085] Figure 3 A schematic diagram of the structure of a device for determining the maximum available distance for strong vibration acceleration recording provided in one embodiment of the present application.

[0086] like Figure 3 As shown, the device 300 for determining the maximum available distance of strong vibration acceleration recording includes:

[0087] An acquisition module 301 is used to acquire strong vibration acceleration records of target type instruments;

[0088] The first determination module 302 is used to perform statistical comparative analysis on the strong vibration acceleration records based on the data quality group to which each record belongs determined in a factual standard manner, and determine a ground peak acceleration threshold for rapid quality classification of data records monitored by such instruments;

[0089] The second determination module 303 is used to construct a mapping relationship between ground peak acceleration and distance attenuation using a seismic attenuation model, bring the ground peak acceleration threshold into the mapping relationship, and determine the maximum available distance for obtaining data monitored by the target type instrument.

[0090] Regarding the device 300 for determining the maximum available distance of strong vibration acceleration recording in the above-mentioned related embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0091] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by any person familiar with the technology within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for determining the maximum available distance of strong vibration acceleration recording, characterized in that: include: Obtain strong vibration acceleration records of target instruments; Based on the data quality group to which each record belongs determined in a factual standard manner, a statistical comparative analysis is performed on the strong vibration acceleration records to determine a ground peak acceleration threshold for rapid quality classification of data records monitored by such instruments; A mapping relationship between ground peak acceleration and distance attenuation is constructed using a seismic attenuation model, the ground peak acceleration threshold is brought into the mapping relationship, and a maximum available distance for obtaining data monitored by the target type instrument is determined; The data quality groups include groups of unusable records and at least one group of usable records; and the statistical comparative analysis process includes: Determining the ground peak acceleration value represented by each of the strong vibration acceleration records; For a group pair consisting of a group of unavailable records and a group of available records in the data quality group, within the value range of the ground peak acceleration value, a trial value of the ground peak acceleration threshold for classification of the group pair is selected by traversing with a preset step length, and a binary classification confidence index corresponding to the trial value is determined based on the following expression: in, represents the binary classification confidence index, A represents the proportion of records correctly classified as the group of available records by the trial value, and R represents the proportion of records correctly classified as the group of unavailable records by the trial value; The values ​​of the binary confidence indexes obtained are compared, and the trial value corresponding to the binary confidence index with the largest value is used as the ground peak acceleration threshold for group classification of the available records.

2. The method according to claim 1, wherein: The process of determining the data quality group to which each record belongs in a de facto standard manner includes: Dividing the time series data represented by the strong vibration acceleration record into signal segments and noise segments; Comparing the Fourier amplitude spectra of the signal segment and the noise segment to obtain a signal-to-noise ratio curve representing the signal-to-noise ratio of each frequency point; According to the signal-to-noise ratio curve, with the signal-to-noise ratio being greater than or equal to a first threshold as a cutoff condition, determining a low-frequency cutoff frequency and a high-frequency cutoff frequency of the strong vibration acceleration record, and obtaining an available frequency band feature of the strong vibration acceleration record; According to the available frequency band characteristics of each strong vibration acceleration record, the data quality group to which the corresponding record belongs is determined.

3. The method according to claim 2, wherein: The first threshold is 3.

4. The method according to claim 1, wherein: The process of selecting the trial value of the ground peak acceleration threshold for group classification by traversing within the value range of the ground peak acceleration value with a preset step length includes: Based on the actual data characteristics recorded in the group pair, intercepting the threshold estimated value range from the value range of the ground peak acceleration value; Within the threshold estimation value range, a trial value of the ground peak acceleration threshold for group classification is selected by traversing with a preset step length.

5. The method according to claim 1, wherein: The mapping relationship between the ground peak acceleration and the distance attenuation is constructed by using the earthquake attenuation model, specifically: The earthquake motion attenuation model is used to construct the mapping relationship between the attenuation of the ground peak acceleration and the fault distance or the source distance in earthquake events of different magnitudes.

6. A device for determining the maximum usable distance of strong vibration acceleration recording, characterized in that: include: An acquisition module is used to acquire strong vibration acceleration records of target instruments; A first determination module is used to perform statistical comparative analysis on the strong vibration acceleration records based on the data quality group to which each record belongs determined in a factual standard manner, and determine a ground peak acceleration threshold for rapid quality classification of data records monitored by such instruments; The second determination module is used to construct a mapping relationship between ground peak acceleration and distance attenuation using a seismic attenuation model, bring the ground peak acceleration threshold into the mapping relationship, and determine the maximum available distance for obtaining data obtained by instrument monitoring of the target type; The data quality groups include groups of unusable records and at least one group of usable records; and the statistical comparative analysis process includes: Determining the ground peak acceleration value represented by each of the strong vibration acceleration records; For a group pair consisting of a group of unavailable records and a group of available records in the data quality group, within the value range of the ground peak acceleration value, a trial value of the ground peak acceleration threshold for classification of the group pair is selected by traversing with a preset step length, and a binary classification confidence index corresponding to the trial value is determined based on the following expression: in, represents the binary classification confidence index, A represents the proportion of records correctly classified as the group of available records by the trial value, and R represents the proportion of records correctly classified as the group of unavailable records by the trial value; The values ​​of the binary confidence indexes obtained are compared, and the trial value corresponding to the binary confidence index with the largest value is used as the ground peak acceleration threshold for group classification of the available records.

Images