A method and system for predicting abnormal data of seismic stations using similar earthquakes

Abstract

The present invention provides a method and system for predicting abnormal data of seismic stations using similar earthquakes, which relates to the field of disaster early warning. The method includes: obtaining the seismic phases of several seismic events monitored by the seismic station, generating seismic data templates based on the seismic station and the seismic phases of each seismic event obtained; determining abnormal stations among several seismic stations to be detected that record similar seismic events in a preset area; classifying the causes of abnormal records of abnormal seismic stations to obtain abnormal classification data; using a logistic regression algorithm to train an abnormal record prediction model based on the seismic data templates of the abnormal seismic stations and the corresponding abnormal classification data; and using the model to predict the abnormal causes of abnormal records of abnormal seismic stations. The present invention first determines abnormal seismic stations through similar seismic events, and then trains an abnormal prediction model based on the abnormal seismic stations to predict abnormal records of the abnormal seismic stations, thereby achieving the purpose of quickly correcting abnormal seismic stations.

Description

Technical Field

[0001] The present invention relates to the technical field of disaster early warning, and in particular to a method and system for predicting abnormal data of seismic stations by using similar earthquakes. Background Art

[0002] Among seismic events recorded by seismic stations, some have waveforms that are highly similar. These events are known in the prior art as similar seismic events. Because the waveforms recorded by seismic stations are the result of the coupling of the event's time-source function, the seismic wave propagation path, and the seismometer instrument response, waveform similarity indicates similar focal mechanisms and close spatial locations. In the prior art, similar seismic events are a highly effective means of studying temporal changes in the Earth's interior.

[0003] To strengthen earthquake monitoring, prediction, and early warning capabilities, a multidisciplinary earthquake monitoring system must be established, with unified planning and management of seismic observation stations to create a high-quality seismic monitoring network. Obtaining complete, continuous, and reliable seismic observation data from seismic stations provides fundamental data for short- and medium-term earthquake predictions within the region, a reliable scientific basis for medium- and short-term earthquake predictions, and post-earthquake trend assessment. Furthermore, it serves regional earthquake preparedness and disaster reduction, improves regional earthquake prediction and forecasting capabilities, and provides fundamental observational data for geoscience research. Seismic stations play a vital role in earthquake prediction, post-earthquake trend assessment, and geoscience research, and their data must be accurate and reliable. If a seismic station is in an abnormal state, the seismic data it continuously monitors may not represent the actual earthquake environment and cannot provide reliable data support. Therefore, it is crucial to ensure that seismic stations are in normal working order.

[0004] Based on the authenticity and efficiency of similar earthquake events as a tool for studying the temporal changes of the Earth's internal medium, researchers are considering using the highly similar waveform characteristics of seismic stations when recording similar earthquake events to measure the recording capability of seismic stations. In other words, the quality of natural earthquake data recorded by a seismic station is evaluated based on its ability to record similar earthquake events. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for predicting abnormal data of seismic stations using similar earthquakes. First, abnormal seismic stations are identified through similar earthquake events, and an abnormal record prediction model is trained based on the abnormal classification data corresponding to the abnormal seismic stations and the abnormal causes. Then, the abnormal causes of the abnormal seismic stations are predicted based on the abnormal record prediction model. The scheme quickly corrects the abnormalities of the abnormal seismic stations and restores them to a normal monitoring state by predicting the abnormal causes of the abnormal seismic stations.

[0006] To achieve the above-mentioned object, the present invention proposes the following technical solution: a method for predicting abnormal data of seismic stations using similar earthquakes, comprising:

[0007] Obtaining seismic phases of a plurality of seismic events monitored by a seismic station, and generating a seismic data template according to the seismic station and the seismic phases of each seismic event obtained;

[0008] Acquire similar earthquake events and determine abnormal earthquake stations among several earthquake stations to be detected that record similar earthquake events in a preset area;

[0009] The abnormal records of abnormal seismic stations are classified by causes to obtain abnormal classification data; the abnormal classification data includes seismograph calibration signal interference, instrument self-noise interference, instrument centroid alignment signal interference, and regional long-period interference;

[0010] Obtain earthquake data templates and corresponding abnormal classification data of abnormal earthquake stations, use logistic regression algorithm to train abnormal record prediction model, and select logistic regression hyperparameters through grid search method;

[0011] The abnormal record prediction model is used to predict the abnormal causes of abnormal records at abnormal seismic stations.

[0012] Furthermore, the obtaining of similar earthquake events includes:

[0013] The time domain waveform cross-correlation method is used to calculate the first cross-correlation coefficient of the seismic data template of any seismic event monitored by the seismic stations.

[0014] For any two seismic stations for which the first mutual correlation coefficient is calculated, determining the magnitude of the first mutual correlation coefficient and the first correlation threshold;

[0015] Any two seismic stations whose first mutual correlation coefficient is not lower than a first correlation threshold are selected, and the seismic events monitored by the seismic stations are similar seismic events.

[0016] Furthermore, the process of determining an abnormal seismic station among a plurality of seismic stations that record similar seismic events in a preset area includes:

[0017] The time domain waveform cross-correlation method is used to calculate the second cross-correlation coefficient of the seismic data template of the seismic station to be detected that monitors similar seismic events;

[0018] determining a magnitude of the calculated second correlation coefficient and a second correlation threshold;

[0019] Seismic stations whose second correlation coefficients are lower than a second correlation threshold are screened, and the screened seismic stations are abnormal seismic stations.

[0020] Furthermore, it also includes: performing energy verification on abnormal seismic stations determined based on similar seismic events;

[0021] The energy verification includes:

[0022] Obtain waveform data of similar earthquake events recorded by abnormal seismic stations;

[0023] After preprocessing the waveform data, the amplitude scaling factor of each channel waveform data of the abnormal seismic station and the variance of the amplitude scaling factor of each channel are calculated using a sliding window algorithm; wherein the preprocessing includes de-meaning, de-linear trending, 0.01-0.05 Hz band-pass filtering and normalization processing in sequence;

[0024] Statistical hypothesis testing methods were used to calculate the confidence interval of the variance of the amplitude scale factor;

[0025] When the variance of the amplitude scale factor of an abnormal seismic station does not fall within the confidence interval of the variance of the amplitude scale factor, the abnormal seismic station is verified as a confirmed abnormal seismic station.

[0026] Furthermore, it also includes:

[0027] The anomaly record prediction model is calibrated based on the real anomaly classification data of anomaly records from abnormal seismic stations.

[0028] A second aspect of the present invention is to disclose a system for predicting abnormal data of seismic stations using similar earthquakes, the system comprising:

[0029] An acquisition and generation module is used to acquire the seismic phases of a plurality of seismic events monitored by a seismic station, and to generate a seismic data template according to the seismic station and the seismic phases of each seismic event acquired;

[0030] An acquisition and judgment module is used to acquire similar earthquake events and judge abnormal earthquake stations among a number of earthquake stations to be detected that record similar earthquake events in a preset area;

[0031] The classification module is used to classify the causes of abnormal records of abnormal seismic stations and obtain abnormal classification data; the abnormal classification data includes seismograph calibration signal interference, instrument self-noise interference, instrument centroid alignment signal interference, and regional long-period interference;

[0032] The model training module is used to obtain the earthquake data templates of abnormal earthquake stations and the corresponding abnormal classification data, train the abnormal record prediction model using the logistic regression algorithm, and select the logistic regression hyperparameters through the grid search method;

[0033] The prediction module is used to predict the abnormal causes of abnormal records of abnormal seismic stations using an abnormal record prediction model.

[0034] Furthermore, the acquisition and judgment module acquires the execution unit of similar earthquake events, including:

[0035] A first calculation unit is used to calculate the first cross-correlation coefficient of the seismic data template of any seismic event monitored by the seismic station by two-to-two using a time domain waveform cross-correlation method;

[0036] A first judgment unit is configured to judge, for any two seismic stations that calculate the first mutual correlation coefficient, the magnitude of the first mutual correlation coefficient and the first correlation threshold;

[0037] The first screening unit is used to screen any two seismic stations whose first mutual correlation coefficient is not lower than a first correlation threshold, and the seismic events monitored by the seismic stations are similar seismic events.

[0038] Furthermore, the execution unit of the acquisition and judgment module for judging an abnormal seismic station among a plurality of seismic stations that record similar seismic events in a preset area includes:

[0039] The second calculation unit is used to calculate the second cross-correlation coefficient of the seismic data template of the seismic station to be detected monitoring similar seismic events by using the time domain waveform cross-correlation method;

[0040] a second judgment unit, configured to judge the magnitude of the calculated second correlation coefficient and a second correlation threshold;

[0041] The second screening unit is used to screen the seismic stations whose second correlation coefficients are lower than the second correlation threshold, and the screened seismic stations are abnormal seismic stations.

[0042] Furthermore, it also includes:

[0043] Energy verification module, used to perform energy verification on abnormal seismic stations judged based on similar seismic events;

[0044] The execution unit of the energy verification module includes:

[0045] an acquisition unit, for acquiring waveform data of similar earthquake events recorded by abnormal earthquake stations;

[0046] a third calculation unit, configured to calculate the amplitude scaling factor of each channel waveform data of the abnormal seismic station and the variance of the amplitude scaling factor of each channel by using a sliding window algorithm after preprocessing the waveform data; wherein the preprocessing includes sequentially performing mean removal, linear trend removal, 0.01-0.05 Hz bandpass filtering, and normalization processing;

[0047] a fourth calculation unit, configured to calculate a confidence interval of the amplitude scale factor variance using a statistical hypothesis testing method;

[0048] The verification unit is used to verify that the abnormal seismic station is a confirmed abnormal seismic station when the variance of the amplitude scale factor of the abnormal seismic station does not fall within the confidence interval of the variance of the amplitude scale factor.

[0049] The third aspect of the present invention discloses an electronic device comprising at least one processor; the processor is coupled to a memory, the memory being used to store one or more computing instructions, wherein the one or more computer instructions are executed by the processor at runtime to implement the steps of the above-mentioned method for predicting abnormal data of seismic stations using similar earthquakes.

[0050] It can be seen from the above technical solutions that the technical solutions of the present invention have the following beneficial effects:

[0051] The present invention discloses a method and system for predicting abnormal data of seismic stations using similar earthquakes. The method includes: obtaining the seismic phases of several seismic events monitored by the seismic station, generating a seismic data template according to the seismic station and the seismic phases of each seismic event; determining abnormal seismic stations among several to-be-detected seismic stations that record similar seismic events in a preset area; classifying the causes of abnormal records of the abnormal seismic stations to obtain abnormal classification data; obtaining the seismic data templates and corresponding abnormal classification data of the abnormal seismic stations, training an abnormal record prediction model using a logistic regression algorithm; and using the model to predict the abnormal causes of abnormal records of abnormal seismic stations. The present invention first determines abnormal seismic stations through similar seismic events, and then trains an abnormal prediction model based on the abnormal seismic stations to predict abnormal records of the abnormal seismic stations, thereby achieving the purpose of quickly correcting abnormal seismic stations and improving the working efficiency of seismic stations.

[0052] The present invention quickly identifies abnormal seismic stations through similar earthquake events, trains an abnormal record prediction model based on the abnormal analysis data of abnormal seismic stations, and quickly predicts the cause of abnormal records of any abnormal seismic station in the region. The method is not limited to the time range monitored by the seismic station and has a wide range of applications.

[0053] It should be appreciated that all combinations of the foregoing concepts, as well as additional concepts described in greater detail below, to the extent such concepts are not mutually inconsistent, can be considered to be part of the inventive subject matter of this disclosure.

[0054] The foregoing and other aspects, embodiments, and features of the present invention will be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the present invention, such as features and / or beneficial effects of the exemplary embodiments, will become apparent from the following description or through practice of specific embodiments according to the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings are not drawn to scale. In the accompanying drawings, each identical or nearly identical component shown in various figures may be represented by the same reference numeral. For clarity, not every component is labeled in every figure. Embodiments of various aspects of the present invention will now be described by way of example with reference to the accompanying drawings, in which:

[0056] Figure 1 A flow chart of a method for predicting abnormal data of seismic stations disclosed in an embodiment of the present application;

[0057] Figure 2 A flowchart of obtaining similar earthquake events disclosed in an embodiment of the present application;

[0058] Figure 3 This is a flow chart for determining abnormal seismic stations disclosed in an embodiment of the present application;

[0059] Figure 4 This is a flow chart of energy verification disclosed in the embodiment of this application;

[0060] Figure 5 is a schematic diagram of an electronic device according to an embodiment of the present application;

[0061] Figure 6 This is a system framework diagram for predicting abnormal data of seismic stations disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be the common meanings understood by people with ordinary skills in the field to which the present invention belongs.

[0063] The words “first”, “second” and similar words used in the patent application specification and claims of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular forms of “a”, “an” or “the” and similar words do not indicate a quantity limitation, but rather indicate the presence of at least one. Words such as “include” or “comprise” mean that the elements or objects appearing before “include” or “comprises” cover the features, wholes, steps, operations, elements and / or components listed after “include” or “comprises”, and do not exclude the existence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections. “Up”, “down”, “left”, “right” and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0064] Seismic stations can obtain complete, continuous, and reliable seismic observation data, which plays an important role in earthquake prediction, post-earthquake trend determination, and earth science research. The reliability of their data needs to be guaranteed. Similar earthquake events indicate similar focal mechanisms and close spatial locations. Their application in earthquake gestation, occurrence, and post-earthquake recovery proves that these events are an effective means of studying the temporal changes of the Earth's internal medium. Therefore, the present invention proposes a method and system for predicting abnormal data from seismic stations using similar earthquakes. Similar earthquake events are used to identify abnormal seismic stations, and then anomaly analysis of these stations is performed to predict the cause of the anomaly.

[0065] The following is a further detailed introduction to the method and system for predicting abnormal data of seismic stations using similar earthquakes disclosed in the present invention in conjunction with the specific embodiments shown in the accompanying drawings.

[0066] Combine Figure 1 As shown, the method disclosed in the embodiment for predicting abnormal data of seismic stations using similar earthquakes specifically includes the following steps:

[0067] Step S102, obtaining the seismic phases of several seismic events monitored by the seismic station, and generating a seismic data template according to the seismic station and the seismic phases of each seismic event; the seismic template data includes waveform data of the seismic event and seismic station data;

[0068] Step S104, obtaining similar earthquake events, and determining abnormal earthquake stations among several to-be-detected earthquake stations that record similar earthquake events in a preset area;

[0069] Since the waveforms of similar earthquake events are very similar, seismic stations in the same earthquake belt should be able to monitor synchronously and have similar waveforms. This step uses similar earthquake events to screen abnormal seismic stations. On the one hand, the highly similar waveform data recorded by the seismic stations can be used to determine whether the earthquake events belong to similar earthquake events. On the other hand, the recording of clearly similar earthquake events by the seismic stations can be used to reflect the data recording quality of the seismic stations, and then reflect whether the seismic stations are in normal working condition.

[0070] Specifically, the process of determining whether an earthquake event belongs to a similar earthquake event by using highly similar waveform data recorded by a seismic station may include: Figure 2 The following steps are shown: Step S202, using the time domain waveform cross-correlation method, calculate the first cross-correlation coefficient of the seismic data template of any seismic event monitored by the seismic stations; Step S204, for any two seismic stations for which the first cross-correlation coefficient is calculated, determine the magnitude of the first cross-correlation coefficient between the two seismic stations and a first correlation threshold, which can be set to 0.9; Step S206, select any two seismic stations whose first cross-correlation coefficient is not less than the first correlation threshold, and the seismic events monitored by the seismic stations are similar seismic events. Optionally, similar seismic events at two seismic stations can be determined by calculating the cross-correlation coefficient of the waveform data recorded by each channel corresponding to the seismic stations.

[0071] As an optional implementation, similar earthquake events may be obtained by looking up known similar earthquake events according to the CENC earthquake catalog; that is, known similar earthquake events may be directly used to screen abnormal earthquake stations.

[0072] Combine Figure 3 As shown, the process of further determining abnormal seismic stations among several seismic stations that record similar seismic events in a preset area based on similar seismic events includes the following steps: Step S302, using the time domain waveform cross-correlation method to calculate the second cross-correlation coefficient of the seismic data template of the seismic station to be detected monitoring similar seismic events; Step S304, determining the size of the calculated second cross-correlation coefficient and the second correlation threshold, and the second correlation threshold can also be set to 0.9; Step S306, screening seismic stations whose second cross-correlation coefficients are lower than the second correlation threshold, and the screened seismic stations are abnormal seismic stations.

[0073] Step S106, classifying the causes of abnormal records of abnormal seismic stations to obtain abnormal classification data; wherein the abnormal classification data includes seismograph calibration signal interference, instrument self-noise interference, instrument centroid alignment signal interference, and regional long-period interference;

[0074] Interference with seismograph calibration signals can manifest as the absence of seismic waveforms recorded on any channel of a seismic station during a given earthquake event. The station itself is intact, and observation of the raw waveforms reveals a clear fluctuation signal before the calibration signal at this anomalous station, while the waveform recording of the seismic event after calibration is unclear. Instrument self-noise interference can manifest as noise with minimal amplitude fluctuations in the raw waveform recording at the anomalous station, with no seismic signal recorded. This indicates that the seismometer at this station was not functioning properly during this period, recording only the instrument's own noise. Instrument center-of-mass alignment interference can manifest as seismic waveforms recorded on any channel of the station, but with strong, long-period interference preceding the seismic signal. Analysis of the station's waveform records confirms that this type of significant long-period interference is consistent with the characteristics of an instrument center-of-mass alignment signal. Regional long-period interference can manifest as low correlation between the horizontal components of two earthquake events recorded at the station and the presence of periodic PSD peaks, consistent with the characteristics of a regional long-period interference signal.

[0075] This step aims to classify and analyze the causes of abnormal records of different abnormal seismic stations, that is, to obtain abnormal classification data; any classification includes waveform records of multiple abnormal seismic stations.

[0076] As an optional implementation, anomaly classification data can also include precision index interference. Precision index interference occurs when a correlation threshold is selected too low during the processing of anomalous seismic station data. When this threshold is lower than the correlation threshold for common similar earthquakes, there is a certain probability of missed detection, meaning that some anomalous stations are not detected. This anomaly classification can be adjusted through data processing.

[0077] Step S108, obtaining earthquake data templates and corresponding abnormal classification data of abnormal earthquake stations, using a logistic regression algorithm to train an abnormal record prediction model, and selecting logistic regression hyperparameters through a grid search method;

[0078] Step S110: using an abnormal record prediction model to predict the abnormal cause of abnormal records at abnormal seismic stations.

[0079] By matching the seismic data templates of abnormal seismic stations with abnormal classification data, seismic data templates with classification labels are obtained to form a training data set; then the data set is trained using a logistic regression algorithm. During the training process, the grid search method in machine learning is used to select appropriate logistic regression algorithm hyperparameters to obtain the model parameters with the best generalization performance. Finally, the abnormal record prediction model based on logistic regression obtained through training is used to process abnormal seismic stations whose abnormal causes need to be analyzed.

[0080] As an optional implementation, the method disclosed in the embodiment for predicting abnormal data of seismic stations using similar earthquakes further includes step S1042, performing energy verification on abnormal seismic stations determined based on similar earthquake events; that is, energy verification is used to further determine whether the abnormal seismic station is a confirmed abnormal seismic station, otherwise, in-depth research is required on the abnormal seismic station. Specifically, the energy verification process is as follows: Figure 4 As shown, the method includes the following steps: step S402, obtaining waveform data of similar earthquake events recorded by abnormal earthquake stations; step S404, after preprocessing the waveform data, calculating the amplitude scaling factor of each channel waveform data of the abnormal earthquake station and the variance of the amplitude scaling factor of each channel by a sliding window algorithm; wherein the preprocessing includes removing the mean, removing the linear trend, 0.01-0.05 Hz band-pass filtering and normalization processing in sequence; step S406, calculating the confidence interval of the variance of the amplitude scaling factor by using a statistical hypothesis testing method; step S408, when the variance of the amplitude scaling factor of the abnormal earthquake station does not fall within the confidence interval of the variance of the amplitude scaling factor, verifying that the abnormal earthquake station is a confirmed abnormal earthquake station.

[0081] As an optional implementation, when the variance of the amplitude scaling factor of the abnormal seismic station falls within the confidence interval of the amplitude scaling factor variance, the abnormal seismic station is recorded as a pending abnormal station; the scheme repeatedly performs energy verification on the pending abnormal station based on other similar seismic events recorded in the preset area until there is a similar seismic event that verifies it as an abnormal seismic station, otherwise the pending abnormal station is re-recorded as a normal seismic station.

[0082] As an optional embodiment, the method for predicting abnormal data of seismic stations using similar earthquakes further includes step S112, calibrating the abnormal record prediction model based on the actual abnormal classification data of abnormal records of abnormal seismic stations. That is, data analysis is performed on the abnormal causes of the predicted abnormal seismic stations to obtain the actual abnormal classification data of the abnormal seismic stations; the seismic data template and the actual abnormal classification data of the predicted abnormal seismic stations are added to the training set of the abnormal record prediction model, and the abnormal record prediction model is further retrained using a logistic regression algorithm, thereby achieving the technical effect of calibrating and updating the abnormal record prediction model.

[0083] Optionally, four colors are used to distinguish abnormal seismic stations on the seismic network. The colors indicate the predicted abnormal classification, which allows for visual management of the abnormal classification of abnormal seismic stations, thereby selectively obtaining seismic stations in normal working condition during earthquake analysis. In addition, for any color-coded abnormal seismic station, the abnormal occurrence time, abnormal waveform record, and abnormal calibration time are recorded to facilitate data tracing.

[0084] The method disclosed in the above embodiment uses similar earthquakes to predict abnormal data of seismic stations. It quickly identifies abnormal seismic stations through similar earthquake events, trains an abnormal record prediction model based on the abnormal analysis data of abnormal seismic stations, and quickly predicts the cause of abnormal records of any abnormal seismic station in the region. It can be widely applicable to support data anomaly analysis of seismic stations in various regions, and can quickly analyze the cause of anomalies without being limited by the recording time of seismic stations, thereby improving the efficiency of fault analysis at seismic stations.

[0085] In an embodiment of the present application, an electronic device is further provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method disclosed in the above embodiment for predicting abnormal data of seismic stations using similar earthquakes is implemented. Taking an electronic device running on a computer as an example, Figure 5 As shown, the electronic device may include one or more (only one is shown in the figure) processors (the processor may include but is not limited to a microprocessor MCU or a programmable logic device FPGA, etc.), a memory for storing data, and a transmission device for communication functions. It will be understood by those skilled in the art that Figure 5 The structure shown is for illustration only and does not limit the structure of the above-mentioned electronic device.

[0086] The above program can be run in the processor, or it can be stored in the memory, that is, in the computer-readable medium, which includes permanent and non-permanent, removable and non-removable media and can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media such as modulated data signals and carrier waves. These computer programs can also be loaded into a computer or other programmable data processing system.

[0087] Device, causing a series of operational steps to be performed on a computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide the instructions for implementing the process Figure 1a process or multiple processes and / or boxes Figure 1 The functional steps specified in a block or multiple blocks may be implemented by different modules corresponding to different method steps.

[0088] In this embodiment, such an apparatus or system is provided. The system can be called a system for predicting abnormal data of seismic stations by using similar earthquakes. Figure 6 As shown, it includes: an acquisition generation module, which is used to obtain the seismic phases of several seismic events monitored by the seismic station, and generate a seismic data template according to the seismic station and the seismic phases of each seismic event obtained; an acquisition judgment module, which is used to obtain similar seismic events and judge the abnormal seismic stations among several seismic stations to be detected that record similar seismic events in a preset area; a classification module, which is used to classify the causes of abnormal records of abnormal seismic stations and obtain abnormal classification data; wherein the abnormal classification data includes seismograph calibration signal interference, instrument self-noise interference, instrument center of mass centering signal interference, and regional long-period interference; a model training module, which is used to obtain the seismic data templates and corresponding abnormal classification data of abnormal seismic stations, use the logistic regression algorithm to train the abnormal record prediction model, and select the logistic regression hyperparameters through the grid search method; a prediction module, which is used to use the abnormal record prediction model to predict the abnormal causes of abnormal records of abnormal seismic stations.

[0089] The system is used to implement the steps of the method for predicting abnormal data of seismic stations using similar earthquakes disclosed in the above embodiment, which have been described and will not be repeated here.

[0090] For example, the acquisition and judgment module acquires an execution unit for similar earthquake events, including: a first calculation unit, used to use the time domain waveform cross-correlation method to calculate the first mutual correlation coefficient of the earthquake data template of any earthquake event monitored by the seismic stations in pairs; a first judgment unit, used to judge the size of the first mutual correlation coefficient and the first correlation threshold of any two seismic stations for which the first mutual correlation coefficient is calculated; a first screening unit, used to screen any two seismic stations whose first mutual correlation coefficient is not lower than the first correlation threshold, and the earthquake events monitored by the seismic stations are similar earthquake events.

[0091] For another example, the acquisition and judgment module determines the execution unit of an abnormal seismic station among several seismic stations that record similar seismic events in a preset area, including: a second calculation unit, used to calculate the second cross-correlation coefficient of the seismic data template of the seismic station to be detected monitoring similar seismic events by using the time domain waveform cross-correlation method; a second judgment unit, used to judge the size of the calculated second cross-correlation coefficient and the second correlation threshold; a second screening unit, used to screen seismic stations whose second cross-correlation coefficient is lower than the second correlation threshold, and the screened seismic stations are abnormal seismic stations.

[0092] For another example, a system for predicting abnormal data of seismic stations using similar earthquakes also includes: an energy verification module for performing energy verification on abnormal seismic stations judged based on similar seismic events; an execution unit of the energy verification module includes: an acquisition unit for acquiring waveform data of similar seismic events recorded by abnormal seismic stations; a third calculation unit for calculating the amplitude scaling factor of each channel waveform data of the abnormal seismic station and the variance of the amplitude scaling factor of each channel by a sliding window algorithm after preprocessing the waveform data; wherein the preprocessing includes de-meaning, de-linear trending, 0.01-0.05Hz band-pass filtering and normalization processing performed in sequence; a fourth calculation unit for calculating the confidence interval of the variance of the amplitude scaling factor by using a statistical hypothesis testing method; and a verification unit for verifying that the abnormal seismic station is a confirmed abnormal seismic station when the variance of the amplitude scaling factor of the abnormal seismic station does not fall within the confidence interval of the variance of the amplitude scaling factor.

[0093] For another example, the system for predicting abnormal data of seismic stations using similar earthquakes further includes a calibration module for calibrating the abnormal record prediction model based on actual abnormal classification data of abnormal records of abnormal seismic stations.

[0094] The above embodiments of the present application disclose a method and system for predicting abnormal data of seismic stations using similar earthquakes. On the one hand, similar earthquake events are used to identify abnormal seismic stations, and similar earthquake events are applied to the quality analysis of seismic station data, which can efficiently identify abnormal seismic stations. On the other hand, a logistic regression algorithm is used to train a prediction model for abnormal records of abnormal seismic stations. When applied, the abnormal classification of seismic stations can be obtained with calibration accuracy, providing direction for correcting the abnormalities of abnormal seismic stations, and ultimately achieving the purpose of improving the working efficiency of seismic stations.

[0095] While the present invention has been disclosed above with reference to preferred embodiments, this is not intended to limit the present invention. Persons skilled in the art will readily appreciate that various modifications and variations can be made without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A method for predicting abnormal data of seismic stations using similar earthquakes, characterized in that: include: Obtaining the seismic phases of several earthquake events monitored by the seismic station, and generating earthquake data templates according to the seismic station and the seismic phases of each earthquake event obtained; Acquire similar earthquake events and determine abnormal earthquake stations among several earthquake stations to be detected that record similar earthquake events in a preset area; The abnormal records of abnormal seismic stations are classified by causes to obtain abnormal classification data; the abnormal classification data includes seismograph calibration signal interference, instrument self-noise interference, instrument centroid alignment signal interference, and regional long-period interference; Obtain earthquake data templates and corresponding abnormal classification data of abnormal earthquake stations, use logistic regression algorithm to train abnormal record prediction model, and select logistic regression hyperparameters through grid search method; The abnormal record prediction model is used to predict the abnormal causes of abnormal records at abnormal seismic stations.

2. The method for predicting abnormal data of seismic stations using similar earthquakes according to claim 1, characterized in that: The obtaining of similar earthquake events includes: The time domain waveform cross-correlation method is used to calculate the first cross-correlation coefficient of the seismic data template of any seismic event monitored by the seismic stations. For any two seismic stations for which the first mutual correlation coefficient is calculated, determining the magnitude of the first mutual correlation coefficient and the first correlation threshold; Any two seismic stations whose first mutual correlation coefficient is not lower than a first correlation threshold are selected, and the seismic events monitored by the seismic stations are similar seismic events.

3. The method for predicting abnormal data of seismic stations using similar earthquakes according to claim 1, characterized in that: The process of determining an abnormal seismic station among a plurality of seismic stations that record similar seismic events in a preset area includes: The time domain waveform cross-correlation method is used to calculate the second cross-correlation coefficient of the seismic data template of the seismic station to be detected that monitors similar seismic events; determining a magnitude of the calculated second correlation coefficient and a second correlation threshold; Seismic stations whose second correlation coefficients are lower than a second correlation threshold are screened, and the screened seismic stations are abnormal seismic stations.

4. The method for predicting abnormal data of seismic stations using similar earthquakes according to claim 1, characterized in that: Also includes: Energy calibration of abnormal seismic stations identified based on similar seismic events; The energy verification includes: Obtain waveform data of similar earthquake events recorded by abnormal seismic stations; After preprocessing the waveform data, the amplitude scaling factor of each channel waveform data of the abnormal seismic station and the variance of the amplitude scaling factor of each channel are calculated using a sliding window algorithm; wherein the preprocessing includes de-meaning, de-linear trending, 0.01-0.05 Hz band-pass filtering and normalization processing in sequence; Statistical hypothesis testing methods were used to calculate the confidence interval of the variance of the amplitude scale factor; When the variance of the amplitude scale factor of an abnormal seismic station does not fall within the confidence interval of the variance of the amplitude scale factor, the abnormal seismic station is verified as a confirmed abnormal seismic station.

5. The method for predicting abnormal data of seismic stations using similar earthquakes according to claim 1, characterized in that: Also includes: The anomaly record prediction model is calibrated based on the real anomaly classification data of anomaly records from abnormal seismic stations.

6. A system for predicting abnormal data of seismic stations using similar earthquakes, characterized in that: include: An acquisition and generation module is used to acquire the seismic phases of a plurality of seismic events monitored by a seismic station, and to generate a seismic data template according to the seismic station and the seismic phases of each seismic event acquired; An acquisition and judgment module is used to acquire similar earthquake events and judge abnormal earthquake stations among a number of earthquake stations to be detected that record similar earthquake events in a preset area; The classification module is used to classify the causes of abnormal records of abnormal seismic stations and obtain abnormal classification data; the abnormal classification data includes seismograph calibration signal interference, instrument self-noise interference, instrument centroid alignment signal interference, and regional long-period interference; The model training module is used to obtain the earthquake data templates of abnormal earthquake stations and the corresponding abnormal classification data, train the abnormal record prediction model using the logistic regression algorithm, and select the logistic regression hyperparameters through the grid search method; The prediction module is used to predict the abnormal causes of abnormal records of abnormal seismic stations using an abnormal record prediction model.

7. The system for predicting abnormal data of seismic stations using similar earthquakes according to claim 6, characterized in that: The acquisition and judgment module acquires an execution unit of similar earthquake events, including: A first calculation unit is used to calculate the first cross-correlation coefficient of the seismic data template of any seismic event monitored by the seismic station by two-to-two using a time domain waveform cross-correlation method; A first judgment unit is configured to judge, for any two seismic stations that calculate the first mutual correlation coefficient, the magnitude of the first mutual correlation coefficient and the first correlation threshold; The first screening unit is used to screen any two seismic stations whose first mutual correlation coefficient is not lower than a first correlation threshold, and the seismic events monitored by the seismic stations are similar seismic events.

8. The system for predicting abnormal data of seismic stations using similar earthquakes according to claim 6, characterized in that: The acquisition and judgment module determines an execution unit of an abnormal seismic station among a plurality of seismic stations that record similar seismic events in a preset area, including: The second calculation unit is used to calculate the second cross-correlation coefficient of the seismic data template of the seismic station to be detected monitoring similar seismic events by using the time domain waveform cross-correlation method; a second judgment unit, configured to judge the magnitude of the calculated second correlation coefficient and a second correlation threshold; The second screening unit is used to screen the seismic stations whose second correlation coefficients are lower than the second correlation threshold, and the screened seismic stations are abnormal seismic stations.

9. The system for predicting abnormal data of seismic stations using similar earthquakes according to claim 6, characterized in that: Also includes: Energy verification module, used to perform energy verification on abnormal seismic stations judged based on similar seismic events; The execution unit of the energy verification module includes: an acquisition unit, for acquiring waveform data of similar earthquake events recorded by abnormal earthquake stations; a third calculation unit, configured to calculate the amplitude scaling factor of each channel waveform data of the abnormal seismic station and the variance of the amplitude scaling factor of each channel by using a sliding window algorithm after preprocessing the waveform data; wherein the preprocessing includes sequentially performing mean removal, linear trend removal, 0.01-0.05 Hz bandpass filtering, and normalization processing; a fourth calculation unit, configured to calculate a confidence interval of the amplitude scale factor variance using a statistical hypothesis testing method; The verification unit is used to verify that the abnormal seismic station is a confirmed abnormal seismic station when the variance of the amplitude scale factor of the abnormal seismic station does not fall within the confidence interval of the variance of the amplitude scale factor.

10. An electronic device, characterized in that: The method comprises at least one processor; the processor is coupled to a memory, the memory is used to store one or more computing instructions, wherein the one or more computer instructions are executed by the processor at runtime to implement the steps of the method for predicting abnormal data of seismic stations using similar earthquakes as described in any one of claims 1 to 5.

Images