Seismometer azimuth angle intelligent monitoring and operation and maintenance management system

Abstract

The invention relates to a seismometer azimuth angle intelligent monitoring and operation and maintenance management system. The system comprises a data automatic access and processing engine, an azimuth angle state intelligent diagnosis and early warning module, a visual interaction and operation and maintenance management platform and an open data service interface. According to the invention, real-time monitoring and accurate correction of azimuth angle deviation can be realized, the accuracy of station data of the whole network is obviously increased, and a reliable data basis is provided for earthquake monitoring and early warning and scientific research work; a large amount of manual troubleshooting workload is automatically reduced, the work order processing period is remarkably shortened, and limited operation and maintenance resources are concentrated on stations which need to be processed on site; a whole-process closed loop of monitoring, diagnosis, early warning, operation and maintenance and verification is established, all operations and data are traced in the whole process, and the quality management and control requirements of the seismic network are met; seamless joint of correction parameters and scientific research software is achieved through an API interface, scientific research errors caused by historical abnormal data are avoided, and meanwhile labeled and long-term data are provided for geophysical research.

Description

Technical Field

[0001] This invention relates to the field of earthquake monitoring technology, and in particular to an intelligent monitoring and maintenance management system for seismometer azimuth angle. Background Technology

[0002] The accuracy of seismometer azimuth directly determines the reliability of seismic observation data and is a core foundation for earthquake monitoring, source parameter inversion, and geophysical research. Current azimuth deviation detection and management models in the industry still have significant limitations, mainly relying on manual post-processing data analysis or semi-automatic processing (e.g., Niu and Li, 2011; Braunmiller et al., 2020). Specific problems and their derivative impacts are as follows:

[0003] 1. Significant lag and high data quality risk: Problem discovery often lags behind the data acquisition cycle by months to years, resulting in a large amount of real-time observation data having compromised reliability due to azimuth deviation. This not only affects the accuracy of real-time earthquake monitoring and early warning, but may also lead to biased scientific research conclusions and even mislead emergency decision-making.

[0004] 2. Lack of systematic monitoring capabilities: The China Earthquake Networks Center has formed a network of nearly a thousand broadband fixed stations. The existing model cannot achieve continuous, automatic, and full-coverage monitoring of the azimuth status of all stations in the network, making it difficult to accurately grasp the overall situation of the azimuth quality of the entire network of seismometers.

[0005] 3. Broken operation and maintenance loop, resulting in low response efficiency: Test results are mostly presented in the form of database records and paper charts, lacking standardized connection processes with on-site calibration, maintenance and other operation and maintenance work. There is an information gap in "problem discovery - problem handling - effect verification", which leads to delayed operation and maintenance response, a lot of repetitive work and serious waste of resources.

[0006] 4. Lack of early warning mechanism and passive fault handling: It is impossible to identify and warn of potential faults such as loose instruments, human error, and aging components in the early stage. It can only intervene passively after the observation deviation exceeds the fault tolerance threshold, missing the best maintenance opportunity and further aggravating the loss of data quality.

[0007] Although existing technologies (including the applicant's published research results) have established azimuth deviation calculation methods and basic databases, an intelligent management system that can directly serve the daily operation and maintenance of the seismic network and realize the full-link connection of "monitoring-diagnosis-early warning-operation and maintenance-verification" has not yet been formed, which is difficult to meet the actual needs of high-quality operation of the seismic network. Summary of the Invention

[0008] This invention provides an intelligent monitoring and operation and maintenance management system for seismometer azimuth angles. This system can solve the problem in the prior art that there is no intelligent management system that can directly serve the daily operation and maintenance of the seismic network and realize the full-link connection of "monitoring-diagnosis-early warning-operation and maintenance-verification", which makes it difficult to meet the actual needs of high-quality operation of the seismic network.

[0009] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: a seismometer azimuth intelligent monitoring and operation and maintenance management system, comprising: Automated data access and processing engine: Retrieves target network data from seismic data sources via standardized interfaces. Based on preset rules for magnitude and epicentral distance, it filters out suitable teleseismic events for calculation and calculates the theoretical inverse azimuth angle using seismic latitude and longitude and station latitude and longitude. The actual reverse azimuth is determined based on the energy minimization criterion of the P-wave (i.e., seismic P-wave) in the tangential component. ,according to , Obtain azimuth deviation value The output standardization results remove outliers from the azimuth deviation values ​​and take the average value as the average azimuth deviation for that period. At the same time, take twice the 95% confidence interval of the average value as the error value of the azimuth deviation for that period. Azimuth status intelligent diagnosis and early warning module: It takes the standardized results and raw event data output by the processing engine as direct input and core criteria, uses a time-series database to store information of each station throughout its entire life cycle and related operation and maintenance records, and integrates multi-dimensional diagnostic logic of threshold judgment, statistical analysis and artificial intelligence prediction to achieve anomaly identification; Visualized Interaction and Operation and Maintenance Management Platform: Based on the GIS geographic information system, with provincial administrative regions as the boundaries, it gathers and displays the azimuth status of all stations and their intelligent diagnostic and early warning information, establishes a unique digital file for each station, and constructs a closed-loop operation and maintenance work order system with a full-process online management of discovery, dispatching, handling, verification and closure; Open Data Service Interface: Based on the dedicated digital archives built on the visual interaction and operation and maintenance management platform, data access capabilities are provided externally through RESTful standardized API interfaces (network service interfaces designed based on the REST architecture style, which realize efficient and scalable communication between clients and servers through unified resource location, operation methods and status code specifications). The interface uses API key authentication (verifying the legitimacy of requests by transmitting a unique key (APIKey) between clients and servers) and HTTPS encrypted transmission (a network protocol that ensures secure data transmission through encryption technology, introducing an SSL / TLS protocol layer on top of HTTP to ensure that the communication content between clients and servers is not stolen or tampered with) to ensure communication security, and limits the access frequency of a single IP address (a unique digital identifier assigned to network devices for accurate location of devices on the Internet and to realize data transmission) to ensure service stability and data security.

[0010] Preferably, the data automation access and processing engine calculates in the following manner : Known earthquake latitude and longitude ( ) and station latitude and longitude ( (Unit: radians), then the theoretical reverse azimuth angle from the station to the epicenter is... The calculation formula is: .

[0011] Preferably, the data automation access and processing engine is determined in the following manner. Within the selected P-wave (i.e., seismic P-wave) time window [t1, t2], the tangential component energy is searched to maximize its efficiency. The minimum θ value is reached, and this value is the actual reverse azimuth angle. .

[0012] Preferably, the data automation access and processing engine removes outliers and takes the average value in the following way: for azimuth deviation values ​​that meet the set threshold, outliers are first removed using the median absolute deviation method, and then random sampling with replacement is performed. The arithmetic mean of 3,000 to 6,000 samples is taken as the average value of the azimuth deviation.

[0013] Preferably, the anomaly identification of the azimuth state intelligent diagnosis and early warning module includes: a. Threshold alarm: A preset dynamic threshold is set. When the absolute value of the current deviation of the station or the absolute value of the average deviation over the past 30 to 60 days exceeds the threshold, an alarm is automatically triggered. b. Sudden Change Detection: The Median Absolute Deviation (MAD) combined with the CUSUM control chart algorithm (a statistical process control tool based on the principle of sequential analysis, which detects small shifts in the mean by accumulating the deviation between process data and the target value) is used to detect sudden changes in the deviation sequence. Even if the threshold is not exceeded after the change, an early warning is still triggered, which is speculated to be due to human or environmental interference. c. Trend prediction: Based on the characteristics of different station deviation sequences, the time series model is adaptively selected according to the linear or nonlinear trend (ARIMA (Autoregressive Integral Moving Average) model is used for linear trends, and LSTM (Long Short-Term Memory) network is used for nonlinear complex trends). Past deviation data is input to predict future trends. If the predicted value reaches 70%~100% of the threshold, an early warning is triggered to achieve predictive maintenance. d. Special pattern recognition: Built-in typical abnormal pattern recognition model, through feature matching and hypothesis testing, automatically identifies special cases such as deviation close to ±180º, E / N component confusion (north and east data are swapped, causing azimuth determination error), and E / N component polarity reversal (the positive and negative poles of a single component signal are reversed, causing the waveform to be mirrored, resulting in completely wrong initial sign and phase information of the seismic phase), and outputs the cause inference with a confidence level of ≥85%.

[0014] Preferably, the exclusive digital archives established by the visualization interaction and operation and maintenance management platform include: a. Time-series trend chart: Visually displays the azimuth deviation change curve since the station joined the network, and marks the operation and maintenance time and the time point of sudden change in deviation value; b. Quality Comparison Chart: Displays a comparison of seismic waveforms, signal-to-noise ratio, and azimuth accuracy before and after correction, providing a visual representation of the maintenance results; c. Operation and maintenance history records: Completely record all alarms, work orders, calibration operations, and verification results information to form a "station full life cycle quality traceability chain".

[0015] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows: 1. Significantly improved data quality: Real-time monitoring and accurate correction of azimuth deviation are achieved, and the confidence level of data from all stations in the network has increased significantly, providing a reliable data foundation for earthquake monitoring and early warning, and scientific research. 2. Revolutionary improvement in operation and maintenance efficiency: Automation reduces a large amount of manual troubleshooting work, significantly shortens the work order processing cycle, and concentrates limited operation and maintenance resources on stations that require on-site handling; 3. Closed-loop management and traceability: Establish a closed-loop process of "monitoring - diagnosis - early warning - operation and maintenance - verification", with all operations and data recorded throughout the process, meeting the quality control and auditing requirements of the seismic network; 4. Deep collaboration between scientific research and operation and maintenance: Through API interface, the calibration parameters are seamlessly connected with scientific research software to avoid scientific research errors caused by historical abnormal data, while providing labeled and long-term data for geophysical research; 5. Strong scalability and adaptability: The system architecture supports network expansion (compatible with adding new stations and data sources), supports algorithm model iteration (can integrate new algorithms such as Rayleigh wave polarization method), and can be extended to monitor other instrument parameters such as gain and frequency response; 6. Significantly reduced operation and maintenance costs: Greatly reduces labor and transportation costs caused by ineffective operation and maintenance and repeated troubleshooting. Attached Figure Description

[0016] Figure 1 This is a diagram of the overall system architecture of the present invention; Figure 2 This is a system layered architecture diagram of the present invention. Detailed Implementation

[0017] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the technical solution of this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0018] like Figures 1 to 2 As shown, the intelligent monitoring and maintenance management system for seismograph azimuth angle includes: Automated data access and processing engine: Retrieves target network data from seismic data sources via standardized interfaces. Based on preset rules for magnitude and epicentral distance, it filters out suitable teleseismic events for calculation and calculates the theoretical inverse azimuth angle using seismic latitude and longitude and station latitude and longitude. The actual reverse azimuth is determined based on the energy minimization criterion of the P-wave (i.e., seismic P-wave) in the tangential component. ,according to , Obtain azimuth deviation value The output standardization results remove outliers from the azimuth deviation values ​​and take the average value as the average azimuth deviation for that period. At the same time, take twice the 95% confidence interval of the average value as the error value of the azimuth deviation for that period. Azimuth status intelligent diagnosis and early warning module: It takes the standardized results and raw event data output by the processing engine as direct input and core criteria, uses a time-series database to store information of each station throughout its entire life cycle and related operation and maintenance records, and integrates multi-dimensional diagnostic logic of threshold judgment, statistical analysis and artificial intelligence prediction to achieve anomaly identification; Visualized Interaction and Operation and Maintenance Management Platform: Based on the GIS geographic information system, with provincial administrative regions as the boundaries, it gathers and displays the azimuth status of all stations and their intelligent diagnostic and early warning information, establishes a unique digital file for each station, and constructs a closed-loop operation and maintenance work order system with a full-process online management of discovery, dispatching, handling, verification and closure; Open data service interface: Based on the dedicated digital archives built on the visual interaction and operation and maintenance management platform, it provides data access capabilities to the outside world through the RESTI standardized API; the interface adopts API key authentication and HTTPS encrypted transmission to ensure communication security, and limits the access frequency of a single IP address to ensure service stability and data security.

[0019] Specifically, data can be obtained from sources such as the National Earthquake Data Backup Center, regional seismic network waveform libraries, the global earthquake event catalog, and the China Earthquake Networks Center earthquake event catalog. Data acquisition can be automated and timed (configurable daily / every 12 hours) to obtain continuous waveform data, event waveform data, and earthquake metadata (magnitude, epicentral distance, time of occurrence, etc.) from the target network (such as China broadband stations). Automatic data format adaptation (SEED, MINISEED, etc.) is supported. Filtering of teleseismic events can be based on preset rules (such as magnitude ≥ 5.5, epicentral distance 30º-90º).

[0020] To establish a theoretical benchmark for evaluating the accuracy of station azimuth angles, preferably, the automated data access and processing engine calculates the following... : Known earthquake latitude and longitude ( ) and station latitude and longitude ( (Unit: radians), then the theoretical reverse azimuth angle from the station to the epicenter is... The calculation formula is: .

[0021] Specifically, due to various factors, the actual north-facing "N" of a seismometer installation may not coincide with geographic true north, thus affecting the theoretical azimuth angle. and actual azimuth With deviation value The original horizontal components are E ( t (Eastward) and N ( t(Northward), assuming the seismometer's northward "N" completely coincides with the geographic northward "N", the rotation angle is... Calculate the radial component after rotation R ( t ) and tangential components T ( t ):

[0022] When the seismometer's north-direction "N" does not completely coincide with the geographic north-direction "N", there is an azimuth deviation. At this point, the actual reverse azimuth angle is If we further base it on the theoretical reverse azimuth angle Rotating the components will affect the radial component. R ( t ) and tangential components T ( t The amplitude and phase of the ) affect a series of subsequent research results.

[0023] To achieve the goal of automatically estimating the actual arrival direction of seismic waves based on observation data, preferably, the automated data access and processing engine determines the following method: Within the selected P-wave time window [t1,t2], the tangential component energy is searched to achieve optimal performance. The minimum θ value is reached, and this value is the actual reverse azimuth angle. .

[0024] Specifically, the theoretical reverse azimuth angle is known. By solving the actual reverse azimuth angle To indirectly obtain the azimuth deviation value Specifically, based on the energy minimization criterion of the P-wave (i.e., seismic P-wave) in the tangential component: within the selected P-wave time window [t1,t2], the energy of the tangential component is minimized by searching... The minimum θ value is reached, and this value is the actual reverse azimuth angle. In practical discrete computation, minimizing this energy is equivalent to minimizing the root mean square (RMS) value of the tangential component within that time window. The RMS value is calculated for each candidate angle. The RMS value of the tangential component sequence is calculated as the objective function:

[0025] , Where N is the number of sampling points within the time window. The calculation is performed with a step size of 1º within the range θ∈[−180º,180º]. Find the initial optimal angle Near the initial optimal angle, the step size is reduced to 0.1º for local search to precisely determine the optimal angle. smallest Finally, only results that meet the following conditions are retained: vertical and radial cross-correlation coefficient CC>0.8, tangential to radial energy ratio 1-T / R>0.5, radial to vertical energy ratio 1-R / Z>-1, signal-to-noise ratio SNR>5, and waveform integrity ≥90%.

[0026] In order to suppress outlier interference and optimize azimuth deviation estimation, the data automated access and processing engine preferably performs outlier removal and average value calculation in the following manner: for azimuth deviation values ​​that meet the set threshold, outliers are first removed using the median absolute deviation method, and then random sampling with replacement is performed. The arithmetic mean of 3000 to 6000 samples is taken as the average value of the azimuth deviation.

[0027] Specifically, for azimuth deviation values ​​within a set threshold within a date range [D1, D2], outliers are first removed using the median absolute deviation method, followed by random sampling with replacement. The arithmetic mean of 3000 to 6000 samples is taken as the average azimuth deviation, and twice the 95% confidence interval of the average value is taken as the error value of the azimuth deviation.

[0028] It should be noted that the system calls the bng_calc_auto() function in the OrientPy program (which is publicly available) to achieve a fast preliminary estimate, and has made localized enhancements on this basis: for units without external network access, it has a built-in China Earthquake Networks Center (CENC) earthquake event catalog parsing interface, which can directly obtain event information from the internal network data service; it supports users to prepare local earthquake catalog files in advance in accordance with the specified format (such as CSV / JSON), and the system can load the file as the event source, completely eliminating the dependence on external network APIs.

[0029] To improve system reliability, enable predictive maintenance, ensure stable equipment operation, and reduce maintenance costs, the anomaly identification of the azimuth state intelligent diagnosis and early warning module preferably includes: a) Threshold alarm; b) Mutation detection; c) Trend prediction; d) Special pattern recognition.

[0030] Specifically, the intelligent diagnostic rule engine integrates multi-dimensional diagnostic logic, including threshold judgment, statistical analysis, and artificial intelligence prediction, to achieve comprehensive and multi-level identification of anomalies. a. Threshold alarm as a first-level alarm: preset dynamic threshold (supports customization according to station type and regional geological conditions, default 10º). When the absolute value of the current deviation of the station or the absolute value of the average deviation over the past 30 to 60 days exceeds the threshold, an alarm will be automatically triggered, prompting "Immediate on-site correction is required". b. Sudden change detection as a secondary alarm: The Median Absolute Deviation (MAD) combined with the CUSUM control chart algorithm is used to detect sudden changes in the deviation sequence (such as a single change ≥8º). Even if the threshold is not exceeded after the change, an alarm is still triggered, which is speculated to be human or environmental interference such as "instrument collision, loose installation". c. Trend prediction as a level 3 alarm: Based on the characteristics of different station deviation sequences, the time series model is adaptively selected (the ARIMA model is used for linear trends, and the LSTM deep learning model is used for nonlinear complex trends). The deviation data of the past 180 days is input to predict the trend of change in the next 30 days. If the predicted value is close to the threshold (such as reaching 70%~100% of the threshold), an early warning is triggered in advance to achieve "predictive maintenance". d. Special pattern recognition: Built-in 4 types of typical abnormal pattern recognition models (single component polarity reversal, single component swapping, double component polarity reversal, double component polarity reversal and swapping). Through feature matching and hypothesis testing, it automatically identifies special cases such as deviation close to ±180º (polarity reversal), E / N component confusion (component swapping), and E / N component polarity reversal, and outputs cause inferences with confidence ≥85% (such as "suspected BHN component polarity reversal" "high probability E / N component wiring swapping").

[0031] To achieve the goals of establishing a digital archive of the entire equipment lifecycle, real-time monitoring of equipment status, historical traceability, and intelligent operation and maintenance decision support, preferably, the exclusive digital archive established by the visual interaction and operation and maintenance management platform includes: a) Time-series trend chart; b) Quality comparison chart; c) Operation and maintenance history.

[0032] Specifically, a) Time-series trend chart: Visually displays the azimuth deviation change curve since the station joined the network, and marks key nodes (maintenance time, deviation value change time point). b. Quality Comparison Chart: Displays a comparison of seismic waveforms, signal-to-noise ratio, azimuth accuracy, and other indicators before and after correction, providing a visual representation of the maintenance results; c. Operation and maintenance history records: Completely record all alarms, work orders, calibration operations, verification results and other information to form a "station full life cycle quality traceability chain".

[0033] Network-wide Status Visualization: The map overview view, based on a GIS geographic information system, uses provincial administrative regions as boundaries to intuitively display the distribution of all stations. Station status is marked by color coding (green: normal, yellow: warning, red: alarm, blue: special mode anomaly), and supports zooming and filtering (by province, station type, status). Clicking on a station allows for quick viewing of core information (current deviation, most recent alarm time, person in charge); Statistical Dashboard: Real-time display of core indicators such as the overall network station health rate, number of alarm stations, work order processing progress, and regional anomaly distribution, supporting daily / weekly / monthly trend analysis to provide data support for management decisions.

[0034] Closed-loop operation and maintenance work order system: a. Work order generation: Supports automatic (triggered by alarm) or manual (created by administrator) generation of work orders. The system automatically fills in the problem description, suggested correction value, historical data reference, station location, and person in charge information. b. Work order flow: Dispatch is made through the web interface, supporting dual notifications via SMS and platform messages. Maintenance personnel can receive tasks in real time, view station historical data, and correction operation guidelines. c. On-site handling: Maintenance personnel upload on-site photos before and after calibration (which must include location markers and instrument status), measured deviation values, and operation records. Location check-in is supported (to verify the authenticity of on-site handling). d. Automatic verification: After the work order is submitted, the system automatically monitors the data of 3-4 subsequent valid teleseismic events at the station, recalculates the deviation value, and if the deviation is ≤10º, it is judged as "verification passed" and the work order is automatically closed; if the deviation is still >10º, the work order is automatically reopened and upgraded to the provincial administrator, prompting "secondary on-site verification is required"; e. Process Traceability: All operations (work order creation, assignment, handling, verification) are recorded with the operator, time, and content, forming an unalterable operation and maintenance log.

[0035] In practical applications: 1. System Deployment Plan 1.1 Architecture Selection: Adopting a microservice architecture + cloud-native deployment, supporting public cloud (such as Alibaba Cloud, Huawei Cloud) or private cloud deployment, core services (data access, computing engine, diagnostic module, platform services) are deployed independently and can be elastically scaled; 1.2 Hardware configuration: Recommended CPU ≥ 32 cores, memory ≥ 64GB, storage ≥ 1TB (time series database should be configured with a separate storage array), supporting distributed node expansion to meet the performance requirements after the scale of the network is expanded; 1.3 Data Synchronization: The system uses a combination of scheduled synchronization and incremental updates with external data sources to ensure data timeliness (delay ≤ 24 hours), while also supporting manual synchronization (such as after a major earthquake event).

[0036] 2. Azimuth deviation calculation process 2.1 Data Acquisition: Every morning at midnight, the system automatically retrieves the previous day's earthquake event catalog and corresponding station waveform data from the data source, and performs format parsing and integrity verification; 2.2 Event Filtering: Automatically filter events according to preset rules (magnitude ≥ 5.5, epicentral distance 30º–90º, waveform integrity ≥ 90%) to generate a list of "station-event" pairs to be calculated; 2.3 Distributed Deviation Calculation: The system calls the distributed computing engine to handle massive computing tasks.

[0037] Task scheduling: The computing engine dynamically splits the "station-event" list into a large number of independent computing tasks, which are then distributed through a central task queue.

[0038] Parallel execution: Multiple elastic computing nodes deployed in the cloud environment retrieve tasks from the queue in parallel. Each node independently executes the complete deviation calculation process (as shown in the GS.QTS station example below), achieving full utilization of computing resources and high-speed parallel processing of tasks.

[0039] Elastic scaling guarantee: The computing engine automatically scales up and down based on real-time load (such as the number of backlogged tasks in the queue and node CPU utilization). During peak data processing periods (such as after a major earthquake), the system automatically scales up to increase the number of computing node instances to quickly process backlogged tasks; during low-load periods, it automatically scales down to save resources, ensuring a balance between processing efficiency and economy.

[0040] 2.4 Quality Control: For the calculation results of each "station-event", strict quality control rules are applied to screen the results, and only valid results that meet the following conditions are retained: vertical and radial cross-correlation coefficient CC>0.8, signal-to-noise ratio SNR>5, tangential and radial energy ratio 1-T / R>0.5, and radial and vertical energy ratio 1-R / Z>-1.

[0041] 3. Example of intelligent early warning and operation and maintenance closed loop 3.1 Alarm Trigger: After the system processes a new teleseismic event, it calculates the average deviation of the GS.QTS station over the past 30 days to be 60.6º±3.5º, which exceeds the default threshold of 10º. This automatically triggers a Level 1 alarm and generates a maintenance work order. 3.2 Work Order Dispatch: The system automatically assigns work orders to the network operation and maintenance manager and notifies them via platform messages and SMS. The work order includes "The station azimuth deviation is 60º, it is recommended to correct it to about 0º on site. Historical data shows that the deviation has been consistently high in the past 3 years." 3.3 On-site handling: After receiving the task, the maintenance personnel navigated to the station and used a laser ruler and a high-precision north finder (error ≤ 0.1º) to measure and confirm that the deviation between the seismometer's north direction and the geographic north direction was 59.8º, which was consistent with the system calculation result. Then, physical correction was performed, and photos before and after correction and measured data were uploaded (the deviation after correction was 0.3º). 3.4 Automatic verification: The system captured 12 valid teleseismic events at the station, and the calculated average deviation was 6.1º±7.8º, which met the "verification passed" condition (≤10º). 3.5 Work order closed loop: The system automatically closes the work order, updates the station status to "normal", and records "On-site calibration completed, verification passed, current deviation 6.1º" in the health record.

[0042] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A seismometer azimuth intelligent monitoring and operation and maintenance management system, characterized in that, include: Automated data access and processing engine: Retrieves target network data from seismic data sources via standardized interfaces. Based on preset rules for magnitude and epicentral distance, it filters out suitable teleseismic events for calculation and calculates the theoretical inverse azimuth angle using seismic latitude and longitude and station latitude and longitude. The actual reverse azimuth angle is determined based on the energy minimization criterion of the seismic P-wave in the tangential component. ,according to , Obtain azimuth deviation value The standardized results are output. Outliers are removed from the azimuth deviation values ​​using the median absolute deviation method. The average value is taken as the average azimuth deviation for that period. At the same time, twice the 95% confidence interval of the average value is taken as the error value of the azimuth deviation for that period. Azimuth status intelligent diagnosis and early warning module: It takes the standardized results and raw event data output by the processing engine as direct input and core criteria, uses a time-series database to store information of each station throughout its entire life cycle and related operation and maintenance records, and integrates multi-dimensional diagnostic logic of threshold judgment, statistical analysis and artificial intelligence prediction to achieve anomaly identification; Visualized Interaction and Operation and Maintenance Management Platform: Based on the GIS geographic information system, with provincial administrative regions as the boundaries, it gathers and displays the azimuth status of all stations and their intelligent diagnostic and early warning information, establishes a unique digital file for each station, and constructs a closed-loop operation and maintenance work order system with a full-process online management of discovery, dispatching, handling, verification and closure; Open data service interface: Based on the dedicated digital archives built on the visual interaction and operation and maintenance management platform, it provides data access capabilities to the outside world through the RESTI standardized API; the interface adopts API key authentication and HTTPS encrypted transmission to ensure communication security, and limits the access frequency of a single IP address to ensure service stability and data security.

2. The intelligent monitoring and maintenance management system for seismograph azimuth angles according to claim 1, characterized in that: The automated data access and processing engine calculates in the following manner : Known earthquake latitude and longitude ( ) and station latitude and longitude ( (Unit: radians), then the theoretical reverse azimuth angle from the station to the epicenter is... The calculation formula is: 。 3. The intelligent monitoring and maintenance management system for seismograph azimuth angle according to claim 1, characterized in that: The data automation access and processing engine is determined in the following manner. Within the selected seismic P-wave time window, the tangential component energy is searched. The minimum θ value is reached, and this value is the actual reverse azimuth angle. .

4. The intelligent monitoring and maintenance management system for seismograph azimuth angle according to any one of claims 1 to 3, characterized in that: The data automation access and processing engine removes outliers and takes the average value in the following way: For azimuth deviation values ​​that meet the set threshold, outliers are first removed using the median absolute deviation method, and then random sampling with replacement is performed. The arithmetic mean of 3,000 to 6,000 samples is taken as the average value of the azimuth deviation.

5. The intelligent monitoring and maintenance management system for seismograph azimuth angle according to claim 1, characterized in that: The anomaly identification of the intelligent azimuth state diagnosis and early warning module includes: a. Threshold alarm: A preset dynamic threshold is set. When the absolute value of the current deviation of the station or the absolute value of the average deviation over the past 30 to 60 days exceeds the threshold, an alarm is automatically triggered. b. Sudden change detection: The median absolute deviation combined with the CUSUM control chart algorithm is used to detect sudden changes in the deviation sequence. Even if the threshold is not exceeded after the change, an early warning is still triggered, which is speculated to be due to human or environmental interference. c. Trend prediction: Based on the characteristics of different station deviation sequences, select a time series model according to linear or nonlinear trends, input past deviation data to predict future trends, and trigger an early warning if the predicted value reaches 70%~100% of the threshold to achieve predictive maintenance. d. Special pattern recognition: The built-in abnormal pattern recognition model automatically identifies special cases such as deviations close to ±180º, E / N component confusion, and E / N component polarity reversal through feature matching and hypothesis testing, and outputs cause inferences with a confidence level of ≥85%.

6. The intelligent monitoring and maintenance management system for seismograph azimuth angle according to claim 1, characterized in that: The dedicated digital archives established by the visualization interaction and operation and maintenance management platform include: a. Time-series trend chart: Visually displays the azimuth deviation change curve since the station joined the network, and marks the operation and maintenance time and the time point of sudden change in deviation value; b. Quality Comparison Chart: Displays a comparison of seismic waveforms, signal-to-noise ratio, and azimuth accuracy before and after correction, providing a visual representation of the maintenance results; c. Operation and maintenance history records: Completely record all alarms, work orders, calibration operations, and verification results information to form a quality traceability chain for the entire life cycle of the station.

Images