Rockburst micro-seismic monitoring and early warning report generation method based on large model agent

Abstract

The invention provides a rockburst micro-seismic monitoring early warning report generation method based on a large model agent, and relates to the technical field of rockburst micro-seismic monitoring, and the method comprises the steps: uploading and initializing intelligent data based on a predefined daily report template and a TBM parameter file; calculating and converting intelligent stake numbers of the dynamic early warning area; real-time data uploading and intelligent integration based on a micro-seismic monitoring sensor; accurate stake number positioning and calculation of the position of the intelligent sensor; uploading and intelligently screening micro-seismic event data; automatically generating a microseismic event spatio-temporal distribution characteristic graph and a microseismic event evolution law graph driven by a large model agent; performing automatic analysis and early warning quantification on the micro-seismic event evolution law diagram by a large model agent; intelligent rock burst grade matching and judgment based on a rule knowledge base; automatic on-site rock burst picture uploading and analysis integration; generating and dynamically updating an intelligent daily report driven by a large model agent; and finally, an accurate and timely daily report with a uniform format is generated.

Description

Technical Field

[0001] The present invention relates to the technical field of rockburst microseismic monitoring, and in particular to a method for generating a rockburst microseismic monitoring early warning report based on a large model intelligent agent. Background Art

[0002] With the expansion of the scale and complexity of deep engineering projects, the management of construction sites has become increasingly important. The traditional method of generating rockburst microseismic monitoring and early warning reports relies on manual data collection and report writing, which is not only time-consuming and labor-intensive, but also prone to errors and difficult to meet the requirements of modern engineering management for efficiency and accuracy. Moreover, rockburst microseismic monitoring and early warning reports involve a lot of professional knowledge and rely heavily on expert experience. In order to improve the level of informatization of construction management, an automated and intelligent method for generating rockburst microseismic monitoring and early warning reports is needed.

[0003] Specifically, in the traditional rockburst microseismic monitoring and early warning report generation method, first, field engineers and technicians need to manually extract parameters from numerous sensors, equipment and records and organize them into daily reports. This process is not only time-consuming and labor-intensive, but also prone to data input errors due to human factors. Secondly, for the intelligent analysis of microseismic events, users often need to rely on other mapping tools to display data, and then manually analyze the results and fill in the daily reports. This method lacks automation support, resulting in low efficiency. Similarly, the interpretation of geological maps and the analysis of rockburst incubation processes rely on the experience and knowledge of experts, which is very unfriendly to novices, making it difficult to accurately analyze the results. Finally, the review and distribution process of daily reports is also mostly manual, which is not only inefficient, but also easily interfered by human factors, limiting the efficiency and reliability of daily report generation. Although some organizations have adopted document editing tools such as Microsoft Word and use predefined templates to standardize report formats, these methods still lack automation functions and require manual input of data and information.

[0004] The existing technology currently has the following defects:

[0005] (1) Heavy reliance on human experience: Advanced functions such as intelligent analysis of microseismic events, intelligent interpretation of geological maps, and analysis of rockburst incubation processes are currently highly dependent on the experience and knowledge of experts, which poses a high threshold for beginners, making it difficult for them to quickly and accurately master and apply these analysis techniques.

[0006] (2) Inability to conduct intelligent analysis and automatically generate daily reports: Manually integrating and updating scattered data sources into rockburst monitoring and early warning reports is a tedious task. The lack of an automated implementation process prolongs the time it takes to compile daily reports and reduces on-site work efficiency.

[0007] In summary, the traditional method of generating rockburst microseismic monitoring and early warning reports is too cumbersome, and an automated and intelligent method of generating rockburst microseismic monitoring and early warning reports is urgently needed to improve the information level of construction management. To this end, a method of generating rockburst microseismic monitoring and early warning reports based on a large model intelligent agent is proposed. This method can automatically collect and analyze data from the construction site, and use a large model intelligent agent to perform microseismic event analysis, geological map interpretation, and rockburst incubation process analysis, thereby generating accurate and reliable rockburst microseismic monitoring and early warning reports. This method not only improves the efficiency and accuracy of generating rockburst microseismic monitoring and early warning reports, but also reduces the risk of human errors, and significantly improves the information level of construction management. Summary of the invention

[0008] In view of the shortcomings of the prior art, the present invention provides a method for generating a rockburst microseismic monitoring and early warning report based on a large model intelligent body. By uploading templates, sensor data, TBM parameter files, microseismic event files and pictures, the large model intelligent body automatically performs tasks such as parameter extraction, table processing, image analysis and rockburst type judgment, and automatically completes the acquisition of daily report parameters, data visualization, generation of picture descriptions and identification of rockburst types. Finally, this information is integrated and filled into the template to generate accurate, timely and uniformly formatted daily reports. This not only improves the efficiency and accuracy of generating rockburst microseismic monitoring and early warning reports, but also reduces the risk of human errors, and significantly improves the informatization level of construction management.

[0009] A method for generating a rockburst microseismic monitoring early warning report based on a large model intelligent agent specifically comprises the following steps:

[0010] Step 1: Intelligent data upload and initialization based on predefined daily templates and TBM parameter files;

[0011] The daily report template is a predefined document framework that specifies the standard format and content of the daily report, including title, number, date, content, conclusion and result graph.

[0012] The TBM parameter file is the operation data of the tunnel boring machine within a set time period, specifically including time, total thrust, penetration, rod chamber pressure of the top shield cylinder, tail mileage, and cutterhead speed;

[0013] Step 2: Intelligent stake number calculation and conversion of dynamic warning area;

[0014] There are two columns of data for the tail mileage in the TBM parameter file. The first column is in meters, and the second column is in millimeters. Convert the first column of data for the tail mileage to kilometers, take the integer part, and add the letter "K" in front of the integer to indicate the kilometer unit; then convert the second column of data for the tail mileage to meters, and round it off. Add the integer after rounding to the corresponding decimal part of the first column data to get the current position stake number;

[0015] The current position stake number is calculated at the set time in the morning and afternoon every day, and the warning distance is set based on the current position stake number, that is, the warning area is obtained;

[0016] Step 3: Real-time data upload and intelligent integration based on microseismic monitoring sensors;

[0017] The microseismic monitoring sensor is set at the construction site, and the real-time data specifically includes the sensor ID, the sensor component X coordinate, Y coordinate and Z coordinate; the intelligent integration is to integrate the real-time data to obtain the microseismic monitoring parameter file;

[0018] Step 4: Accurate stake number positioning and calculation of intelligent sensor position;

[0019] Calculate the average value of the horizontal coordinates of the first four sensors in the microseismic monitoring parameter file. Specifically, add the horizontal coordinates of the first four sensors and divide them by four. Then convert the average value into the pile number format to obtain the specific position of the first sensor. Similarly, calculate the average value of the horizontal coordinates of the last four sensors in the microseismic monitoring parameter file to determine the specific position of the second sensor.

[0020] Step 5: Upload and intelligently filter microseismic event data;

[0021] The microseismic event data comes from sensors deployed at the construction site, and specifically includes microseismic event time, microseismic real event time, microseismic event X-axis coordinate, microseismic event Y-axis coordinate, microseismic event Z-axis coordinate, energy ratio and signal type.

[0022] Step 6: Automatically generate the spatiotemporal distribution characteristic map of microseismic events and the evolution law map of microseismic events driven by the large model intelligent agent;

[0023] The temporal and spatial distribution characteristic diagram of microseismic events includes a distribution characteristic diagram of microseismic events along the tunnel axis and a distribution characteristic diagram of microseismic events along the tunnel cross section. The specific drawing process is as follows:

[0024] The large model intelligent agent automatically screens the real microseismic events in the microseismic event data, and uses the X-axis, Y-axis, Z-axis coordinates and energy ratio of the selected microseismic events to draw a distribution characteristic diagram of the microseismic events along the tunnel axis, which is used to represent the density and distribution law of microseismic events at different positions along the tunnel axis. Specifically, spheres of different colors are used to represent events at different time points, where the volume of the sphere is proportional to the energy ratio. The larger the energy ratio, the larger the volume of the sphere.

[0025] Draw a distribution characteristic map of microseismic events along the tunnel cross section; used to show the density and distribution law of microseismic events in the tunnel lateral direction;

[0026] The specific drawing process of the microseismic event evolution law diagram is as follows: count the real time and energy ratio of microseismic events, calculate the cumulative number of microseismic events and the cumulative microseismic release energy; and use a double-axis line graph to represent it;

[0027] Step 7: The large model intelligent agent automatically analyzes the evolution law diagram of microseismic events and quantifies the early warning;

[0028] The large model agent decomposes the microseismic event evolution law graph into two single line graphs, namely, the cumulative microseismic event number line graph and the cumulative microseismic release energy line graph, and analyzes the cumulative number of microseismic events generated at all times and the cumulative microseismic release energy generated at all times respectively; calculates and obtains the cumulative number of microseismic events generated in the warning area and the cumulative energy released by these microseismic events;

[0029] Step 8: Intelligent matching and judgment of rockburst level based on rule knowledge base;

[0030] After obtaining the cumulative number of microseismic events generated in the warning area and the cumulative energy released by these microseismic events, the large model agent matches these features with the rockburst type standards preset in the rule knowledge base, which contains feature descriptions and identification rules for different rockburst types; according to these rules, the microseismic event features in the image description are analyzed, the closest rockburst type is determined, and the judgment result is output;

[0031] Step 9: Automated on-site rockburst image upload and analysis integration;

[0032] If no rockburst occurred between the previous daily report and the current daily report, there is no need to upload the on-site rockburst pictures, and step 10 can be performed directly; otherwise, upload the on-site rockburst pictures that occurred between the previous daily report and the current daily report; specifically, first obtain the name of the on-site rockburst picture, which is specifically "warning area, rockburst location, rockburst level, rockburst depth", and then obtain the warning time and area, that is, "the recording time of the previous daily report, the warning area of ​​the previous daily report", the excavation time, that is, "the current position of the previous daily report ~ the current position of this daily report, the time of the previous daily report ~ the time of this daily report", the warning level, that is, "the rockburst warning level of the previous daily report", the actual occurrence, that is, "the name of the picture", the construction situation, that is, "the current position of this daily report ~ the current position of the previous daily report" and comparative analysis, that is, "whether the actual occurrence is consistent with the warning level".

[0033] Step 10: Generate and dynamically update intelligent daily reports driven by large model agents;

[0034] The number of the daily report is determined in a cumulative manner, that is, each daily report number is increased by 1 based on the previous one. The recording cycle of the daily report is updated every 12 hours. Each daily report contains a record of the rock burst warning situation and a verification of the warning situation in the previous daily report. The rock burst warning situation specifically includes the warning area, analysis and description of microseismic activity characteristics, and rock burst warning level;

[0035] The daily report specifically includes a first table, the header of the first table is the warning area information, and the specific location information of the sensor provided by the microseismic monitoring parameter file is also stored in the first table;

[0036] The first table also includes a microseismic event spatiotemporal distribution characteristic diagram and a microseismic event evolution law diagram, which shows the distribution and characteristics of rock burst and microseismic activities; analyzing the characteristics of microseismic activities, matching and determining the level of rock burst warning according to the picture description and the information in the knowledge base, and storing it in the first table;

[0037] If no rockburst occurred during the previous daily report, the daily report only contains the first table; otherwise, the second table is added;

[0038] The second table is used to record and analyze the information of rock burst events; the header of the second table uses the pile number segment of the first table recorded in the previous daily report, and the time recorded in the previous daily report will also be added to the second table as a reference for the warning time and area of ​​this rock burst; the record of excavation time will be based on the pile number from the last record to the pile number recorded this time, and the corresponding time. The warning level will be determined according to the rock burst category judgment result in the previous daily report. The actual rock burst event will be recorded by the name of the picture uploaded by the user, and the construction situation will be reflected by the difference in pile numbers from the previous daily report to the current daily report. The comparative analysis will compare the actual rock burst event with the warning result in the previous daily report to determine the accuracy of the warning and the consistency of the actual occurrence; the second table also contains the on-site rock burst pictures uploaded by the user;

[0039] The rockburst microseismic monitoring and early warning method based on the large model intelligent agent is implemented based on the rock mechanics knowledge interactive system, and specifically includes: a data acquisition module, a data processing module, a rule knowledge base module and a daily report generation module;

[0040] The data processing module is used to receive the daily report template, TBM parameter file, real-time data of microseismic monitoring sensors, and microseismic event data, obtain the current pile number, warning area, sensor location, and transmit them to the data processing module;

[0041] The data processing module draws the spatiotemporal distribution characteristic diagram of microseismic events and the evolution law diagram of microseismic events through the large model intelligent body, and performs image analysis and summary through the multimodal large model, and at the same time calculates the cumulative number of microseismic events generated in the warning area and the cumulative energy released by these microseismic events; the rule knowledge base module determines the closest rock burst type through the rule knowledge base module, and outputs the judgment result to the daily report generation module; the daily report generation module generates daily reports through the large model intelligent body.

[0042] The beneficial effects of adopting the above technical solution are:

[0043] The present invention provides a rockburst microseismic monitoring and early warning report generation method based on a large model intelligent body. Through the intelligent rockburst microseismic monitoring and early warning report generation method, real-time monitoring and precise management of the construction site are realized. Through the efficient data processing and analysis of the large model intelligent body, timely and accurate decision-making information is provided to the construction team and management, thereby optimizing the construction monitoring process, improving the safety management level, reducing construction risks, and ensuring the standardization and traceability of construction information, ultimately improving construction management efficiency and project quality.

[0044] The present invention realizes a fully automated process from data collection to daily report compilation, which not only significantly improves the speed and accuracy of information processing, but also greatly reduces the reliance on manual operations, reducing time and cost. The intelligent agent can automatically extract key parameters such as rockburst warning areas and sensor locations, and use the understanding of the large model to conduct in-depth analysis of on-site pictures and accurately identify rockburst levels, thereby providing a more reliable guarantee for construction safety.

[0045] The rockburst microseismic monitoring and early warning report based on a large model intelligent agent generated by the present invention is standardized, structured, easy to understand and operate, enabling the construction team to quickly respond to various situations and adjust the construction plan and safety measures in a timely manner. The standardization of daily reports also promotes the rapid circulation and sharing of information between the construction team, management and regulatory agencies, strengthens communication and collaboration among all parties, and improves the transparency and management efficiency of construction projects. This information sharing mechanism helps to discover and solve problems in a timely manner, and provides comprehensive and accurate data support for construction decisions.

[0046] In general, the present invention not only improves the efficiency and accuracy of construction project management through the application of intelligent technology, but also provides a more powerful guarantee for construction safety. It promotes the modernization and intelligent development of the construction industry, provides a new solution for information management of construction sites, heralds a major change in the construction management method, and leads the industry to develop in a more efficient, safer and smarter direction. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of a method for generating a rockburst microseismic monitoring and early warning report in an embodiment of the present invention;

[0048] Figure 2 This is a flow chart for obtaining rockburst warning levels in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0050] A method for generating rockburst microseismic monitoring and early warning reports based on a large model agent, such as Figure 1 As shown, the specific steps include:

[0051] Step 1: Intelligent data upload and initialization based on predefined daily templates and TBM parameter files;

[0052] The daily report template is a predefined document framework that specifies the standard format and content of the daily report, including title, number, date, content, conclusion and result graph.

[0053] The TBM parameter file is the operation data of the tunnel boring machine within a set time period, including time, total thrust, penetration, rod chamber pressure of the top shield cylinder, tail mileage, and cutterhead speed. These data are crucial for evaluating construction progress and equipment status.

[0054] Step 2: Intelligent stake number calculation and conversion of dynamic warning area;

[0055] The TBM parameter file is used as a key data source to record the operation data of the tunnel boring machine in a specific time period in detail. Among them, "machine tail mileage" is the core indicator for evaluating the progress of tunneling. In order to convert the mileage data into a format that is easier to understand and use, we adopted a specific conversion method: the machine tail mileage in the TBM parameter file has two columns of data, the first column of data is in meters, and the second column of data is in millimeters. The first column of data of the machine tail mileage is converted to kilometers, and the integer part is taken. The letter "K" is added in front of the integer to indicate the kilometer unit; then the second column of data of the machine tail mileage is converted to meters, and rounded to the nearest integer. The integer after rounding is added to the corresponding decimal part of the first column of data to get the current position pile number; for example, if the first machine tail mileage is 237071 and the second is 863, after conversion and calculation, we get the pile number "K237+072".

[0056] The current position pile number is calculated at the set time in the morning and afternoon every day, and the warning distance is set based on the current position pile number, that is, the warning area is obtained; in order to adapt to the frequent updates of daily reports in actual construction, a pile number calculation method based on daily report time is designed in this embodiment. Every day, we will generate two daily reports at ten o'clock in the morning and ten o'clock in the evening. According to the time points of these two daily reports, we can calculate the current position pile number of the tunnel boring machine. Furthermore, in order to enhance the safety and predictability during the construction process, we propose to shorten 25 meters and increase 15 meters on the basis of the current pile number to construct a warning area. The setting of this warning area is intended to provide a buffer distance for the construction team to promptly discover and respond to potential geological changes or other construction risks during the excavation process.

[0057] Step 3: Real-time data upload and intelligent integration based on microseismic monitoring sensors;

[0058] The microseismic monitoring sensor is set at the construction site, and the real-time data specifically includes the sensor ID, the sensor component X coordinate, Y coordinate and Z coordinate; the intelligent integration is to integrate the real-time data to obtain the microseismic monitoring parameter file;

[0059] Step 4: Accurate stake number positioning and calculation of intelligent sensor position;

[0060] Extract key coordinate data from the microseismic monitoring parameter file. This file records the "sensor component X coordinate" of each sensor in detail, providing us with accurate geolocation information. In order to determine the relative position of the sensor, first calculate the average value of the horizontal coordinates of the first four sensors in the microseismic monitoring parameter file. Specifically, add the horizontal coordinates of the first four sensors and divide them by four to get a representative average coordinate value. This value not only reflects the relative position of these sensors in space, but also serves as the basis for our subsequent data processing and analysis. Then convert the average value into the pile number format to obtain the specific position of the first sensor; similarly calculate the average value of the horizontal coordinates of the last four sensors in the microseismic monitoring parameter file to determine the specific position of the second sensor. This meticulous data processing not only ensures that the installation location of the sensor meets the design requirements, but also provides an accurate reference for subsequent engineering implementation and maintenance, thereby optimizing the layout and operation efficiency of the entire system.

[0061] Step 5: Upload and intelligently filter microseismic event data;

[0062] The microseismic event data comes from sensors deployed at the construction site, and specifically includes microseismic event time, microseismic real event time, microseismic event X-axis coordinate, microseismic event Y-axis coordinate, microseismic event Z-axis coordinate, energy ratio and signal type.

[0063] Step 6: Automatically generate the spatiotemporal distribution characteristic map of microseismic events and the evolution law map of microseismic events driven by the large model intelligent agent;

[0064] The temporal and spatial distribution characteristic diagram of microseismic events includes a distribution characteristic diagram of microseismic events along the tunnel axis and a distribution characteristic diagram of microseismic events along the tunnel cross section. The specific drawing process is as follows:

[0065] The spatiotemporal distribution characteristic diagram of microseismic events includes a distribution characteristic diagram of microseismic events along the tunnel axis and a distribution characteristic diagram of microseismic events along the tunnel cross section. The specific drawing process is as follows: the large model agent uses a table processing tool to filter out the microseismic events that actually occurred in the file, and the signal type is 2 or 12, because these types represent actual geological activities rather than background noise. The large model agent uses the X-axis, Y-axis, and Z-axis coordinates and energy ratio of the selected microseismic events to automatically generate a distribution characteristic diagram of microseismic events along the tunnel axis, which is used to represent the density and distribution law of microseismic events at different locations along the tunnel axis. Specifically, different colored spheres are used to represent events at different time points, where the volume of the sphere is proportional to the energy ratio. The larger the energy ratio, the larger the volume of the sphere, which helps us intuitively identify the hotspot area of ​​energy release.

[0066] Draw a distribution characteristic map of microseismic events along the tunnel cross section; it is used to show the density and distribution law of microseismic events in the tunnel lateral direction, so that we can understand the distribution of events in different cross sections. This analysis is of great significance for understanding the stability of tunnel structures and possible geological risks.

[0067] The specific drawing process of the microseismic event evolution law diagram is as follows: In order to deeply analyze the evolution law of microseismic events, the large model intelligent agent will automatically count the time and energy ratio of the real microseismic events, calculate the cumulative number of microseismic events and the cumulative microseismic release energy; because the vertical coordinate value range of these statistical results may be inconsistent, a dual-axis line chart is used to represent them. This kind of chart can show two different measurements at the same time, allowing us to observe the development trend of microseismic events through the changes in the time series. In this way, we can identify the time evolution law and potential periodic changes of microseismic events, which is of great value for the subsequent analysis of image data prediction and prevention of possible geological disasters.

[0068] Step 7: The large model intelligent agent automatically analyzes the evolution law diagram of microseismic events and quantifies the early warning;

[0069] After the previous step, we obtained a line graph of the evolution of microseismic events, which records in detail the trend of microseismic events over time. In order to obtain more accurate analysis results, such as Figure 2 As shown, the large model agent decomposes the microseismic event evolution law graph into two single line graphs, namely, the cumulative microseismic event number line graph and the cumulative microseismic release energy line graph, and analyzes the cumulative number of microseismic events and the cumulative microseismic release energy at all times through the multimodal large model GLM4V respectively; in this embodiment, the two line graphs are analyzed in depth using different prompts "Please analyze the cumulative number of microseismic events at all times in this picture" and "Please analyze the cumulative microseismic release energy at all times in this picture", and the cumulative number of microseismic events in the warning area and the cumulative energy released by these microseismic events are calculated and obtained, for example: "The warning area has accumulated 88 microseismic events and a cumulative microseismic release energy of 32.2J." This analysis is crucial for predicting potential geological risks and formulating corresponding safety measures, because it can help relevant personnel understand the intensity and frequency of microseismic activities, so as to take preventive measures to reduce possible damage.

[0070] Step 8: Intelligent matching and judgment of rockburst level based on rule knowledge base, such as Figure 2 As shown;

[0071] After obtaining the cumulative number of microseismic events generated in the warning area and the cumulative energy released by these microseismic events, the large model agent matches these features with the rockburst type standards preset in the rule knowledge base, such as Figure 2As shown. The rule knowledge base is a database that integrates geological, engineering and historical data. It contains feature descriptions and identification rules for different rockburst types. According to these rules, the daily intelligent agent is generated to analyze the microseismic event features in the picture description, match the most similar fragments as prompt content, and input the prompt content and the question into the large language model for answering. In order to judge the closest rockburst level without outputting other redundant content, it is also necessary to set a prompt "Please judge the level of rockburst based on the amount of accumulated microseismic release energy. Please only output the rockburst level and do not output other content" to limit it. Finally, the judgment result is obtained; provide a scientific basis for engineering decision-making. At the same time, it can assist engineers and geological experts to timely identify and respond to potential rockburst risks before or during construction, thereby improving construction safety and efficiency.

[0072] Step 9: Automated on-site rockburst image upload and analysis integration;

[0073] If no rockburst occurs between the last daily report and the current daily report, there is no need to upload the on-site rockburst pictures, and step 10 can be performed directly; otherwise, upload the on-site rockburst pictures that occurred between the last daily report and the current daily report; specifically, first obtain the name of the on-site rockburst picture, which is specifically "warning area, rockburst location, rockburst level, rockburst depth", for example, "K130+728.5~730.4 area, 9.5~11 o'clock, slight rockburst, depth 0.28m", and then obtain the warning time and area, that is, "the recording time of the last daily report, the warning area of ​​the last daily report", the excavation time, that is, "the current position of the last daily report~the current position of this daily report, the time of the last daily report~the time of this daily report", the warning level, that is, "the rockburst warning level of the last daily report", the actual occurrence, that is, "the name of the picture", the construction situation, that is, "the current position of this daily report~the current position of the last daily report" and the comparative analysis, that is, "whether the actual occurrence is consistent with the warning level".

[0074] Step 10: Generate and dynamically update intelligent daily reports driven by large model agents;

[0075] The numbering of the daily report is determined in a cumulative manner, that is, each time the daily report is generated, the number will be increased by 1 based on the previous one. This numbering mechanism helps to track and manage the continuity of the daily reports. The recording cycle of the daily report is updated every 12 hours to ensure the real-time and accuracy of the information. Each daily report contains a record of the rock burst warning situation and a verification of the warning situation in the previous daily report, which helps to evaluate the accuracy and efficiency of the warning system. The rock burst warning situation specifically includes the warning area, analysis and description of the characteristics of microseismic activity, and the rock burst warning level;

[0076] The daily report specifically includes a first table, the header of which is the warning area information, which is also stored in the first table, and the specific location information of the sensor provided by the microseismic monitoring parameter file is also stored in the first table; the TBM (tunnel boring machine) parameter file is an important part of the daily report template, which provides key information of the warning area and is used as the header of the first table in the daily report template; the data of the warning area is also stored in the first table for quick access and analysis. The microseismic monitoring parameter file provides the specific location information of the sensor, which will also be recorded in the first table to provide basic data for monitoring and analyzing microseismic activities.

[0077] The first table also includes a microseismic event spatiotemporal distribution characteristic diagram and a microseismic event evolution law diagram, which show the distribution and characteristics of rock bursts and microseismic activities. In order to have a deeper understanding of the information represented by these pictures, we obtained descriptions of the pictures through advanced large model technology. These descriptions analyze the characteristics of microseismic activities in detail and are stored in the daily report as part of the analysis and description of microseismic activity characteristics.

[0078] The daily report will also match and determine the level of rockburst warning based on the image description and information in the knowledge base. This step is a key link in the daily report because it directly affects the subsequent response measures and safety prevention plans. The determination of the warning level is based on the assessment of the possibility of rockburst and the comprehensive analysis of the on-site situation, ensuring the scientificity and practicality of the warning.

[0079] During the recording process, the compilation of the daily report follows strict logic and format. If no rockburst occurred from the previous daily report to this daily report, the daily report only contains the first table; otherwise, the second table is added; this can avoid unnecessary redundant information and ensure the simplicity and pertinence of the report. However, once a rockburst occurs, the compilation of the daily report will become more detailed and complicated.

[0080] The second table is used to record and analyze the information of rock burst events; the header of the second table uses the pile number segment of the first table recorded in the previous daily report, and the time recorded in the previous daily report will also be added to the second table as a reference for the warning time and area of ​​this rock burst; the record of excavation time will be based on the pile number from the last record to the pile number recorded this time, as well as the corresponding time, which helps to track the specific time and location of rock burst. The warning level will be determined based on the rock burst category judgment result in the previous daily report, which provides a quantitative assessment of the severity and urgency of rock burst. The actual rock burst event will be recorded by the name of the picture uploaded by the user, which provides specific evidence of the rock burst. The construction situation is reflected by the difference in pile numbers from the previous daily report to the current daily report, which helps to understand the construction progress and the impact of rock burst events on the construction progress. Comparative analysis is an important part of the daily report, which compares the actual rock burst event with the warning result in the previous daily report to determine the accuracy of the warning and the consistency of the actual occurrence; this not only helps to evaluate the effectiveness of the warning system, but also provides a basis for improvement of future warnings. The second table also contains on-site rockburst pictures uploaded by users; these pictures intuitively show the actual situation at the rockburst site and provide important visual information for further analysis and decision-making.

[0081] In summary, the preparation of daily reports is a comprehensive process that integrates data recording, early warning analysis, event verification and on-site evidence presentation. It not only ensures the continuity and accuracy of rockburst monitoring, but also provides important support for construction safety management and risk assessment.

[0082] The rockburst microseismic monitoring and early warning method based on the large model intelligent agent is implemented based on the rock mechanics knowledge interactive system, and specifically includes: a data acquisition module, a data processing module, a rule knowledge base module and a daily report generation module;

[0083] The data processing module is used to receive the daily report template, TBM parameter file, real-time data of microseismic monitoring sensors, and microseismic event data, obtain the current pile number, warning area, sensor location, and transmit them to the data processing module;

[0084] The data processing module draws a microseismic event spatiotemporal distribution feature map and a microseismic event evolution law map through a large model intelligent agent, and performs image analysis and summary through a multimodal large model, and simultaneously calculates the cumulative number of microseismic events generated in the warning area and the cumulative energy released by these microseismic events; the rule knowledge base module determines the closest rock burst type and outputs the determination result to the daily report generation module; the daily report generation module generates daily reports through a large model intelligent agent; wherein the large model intelligent agent is an intelligent computing entity that can more accurately perceive the environment, react and judge, and form and execute decisions;

[0085] The technical effect of this embodiment is:

[0086] (1) Intelligent analysis of microseismic events: It can automatically screen and process microseismic event data and generate a variety of analysis charts, including spatiotemporal distribution feature charts and evolution law charts. This automated process reduces errors caused by human intervention and significantly improves the efficiency of data processing and analysis, making the monitoring and risk assessment of geological activities more accurate.

[0087] (2) Intelligent chart analysis capability: The large model agent uses its intelligent chart analysis capability to match the accumulated data of microseismic events with the rockburst grade standards in the rule knowledge base, and generates analysis charts such as dual-axis line charts to deeply analyze the evolution of microseismic events. This precise analysis method improves the ability to predict potential geological disasters, helps relevant personnel identify risks early and take preventive measures, thereby effectively reducing safety hazards in engineering operations.

[0088] (3) Automatically generate daily reports: The extracted parameter information and analysis conclusions are inserted into the daily report template according to the specified location to generate a structured and standardized daily report, ensuring the consistency and professionalism of the daily report. The automatic generation of daily reports not only greatly improves the efficiency of daily report compilation and ensures the accuracy and consistency of information, but also reduces the burden on staff, allowing them to focus on more valuable tasks.

[0089] The design of the entire system emphasizes automation, intelligence and efficiency. It realizes the automation of the entire process of daily report generation through the independent operation of a single intelligent body. This design not only improves the speed and quality of daily report generation, but also provides a fast, accurate and comprehensive information management solution for the construction site. Through this solution, the construction management process has been optimized, and the quality and safety level of the project have been improved.

[0090] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) to form a technical solution.

Claims

1. A method for generating rockburst microseismic monitoring and early warning reports based on a large model intelligent agent, characterized in that: The following steps are involved: Step 1: Intelligent data upload and initialization based on predefined daily templates and TBM parameter files; Step 2: Intelligent stake number calculation and conversion of dynamic warning area; Step 3: Real-time data upload and intelligent integration based on microseismic monitoring sensors; The microseismic monitoring sensor is set at the construction site, and the real-time data specifically includes the sensor ID, the sensor component X coordinate, Y coordinate and Z coordinate; the intelligent integration is to integrate the real-time data to obtain the microseismic monitoring parameter file; Step 4: Accurate stake number positioning and calculation of intelligent sensor position; Step 5: Upload and intelligently filter microseismic event data; The microseismic event data comes from sensors deployed at the construction site, and specifically includes microseismic event time, microseismic real event time, microseismic event X-axis coordinate, microseismic event Y-axis coordinate, microseismic event Z-axis coordinate, energy ratio and signal type; Step 6: Automatically generate the spatiotemporal distribution characteristic map of microseismic events and the evolution law map of microseismic events driven by the large model intelligent agent; Step 7: The large model intelligent agent automatically analyzes the evolution law diagram of microseismic events and quantifies the early warning; Step 8: Intelligent matching and judgment of rockburst level based on rule knowledge base; After obtaining the cumulative number of microseismic events generated in the warning area and the cumulative energy released by these microseismic events, the large model agent matches these features with the rockburst type standards preset in the rule knowledge base; Step 9: Automated on-site rockburst image upload and analysis integration; Step 10: Generate and dynamically update intelligent daily reports driven by large model agents; The number of the daily report is determined in a cumulative manner, that is, each generated daily report number will be increased by 1 based on the previous one; the recording cycle of the daily report is updated every 12 hours, and each daily report contains a record of the rock burst warning situation and a verification of the warning situation in the previous daily report, wherein the rock burst warning situation specifically includes the warning area, analysis and description of microseismic activity characteristics, and rock burst warning level.

2. According to the method for generating a rockburst microseismic monitoring and early warning report based on a large model intelligent agent according to claim 1, it is characterized in that: The daily report template described in step 1 is a predefined document framework that specifies the standard format and content of the daily report, including title, number, date, content, conclusion and result graph; The TBM parameter file is the operating data of the tunnel boring machine within a set time period, specifically including time, total thrust, penetration, rod chamber pressure of the top shield cylinder, tail mileage, and cutterhead speed.

3. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1 is characterized in that: The step 2 is specifically as follows: the tail mileage in the TBM parameter file has two columns of data, the first column of data is in meters, and the second column of data is in millimeters. The first column of data of the tail mileage is converted to kilometers, and the integer part is taken, and the letter "K" is added in front of the integer to indicate the kilometer unit; then the second column of data of the tail mileage is converted to meters, and rounded to the nearest integer. The integer after rounding is added to the corresponding decimal part of the first column of data to obtain the current position stake number; The current position stake number is calculated at the set time in the morning and afternoon every day, and the warning distance is set based on the current position stake number, that is, the warning area is obtained.

4. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1 is characterized in that: Step 4 is specifically as follows: Calculate the average value of the horizontal coordinates of the first four sensors in the microseismic monitoring parameter file, specifically add the horizontal coordinates of the first four sensors and divide them by four, and then convert the average value into the pile number format to obtain the specific position of the first sensor; similarly calculate the average value of the horizontal coordinates of the last four sensors in the microseismic monitoring parameter file to determine the specific position of the second sensor.

5. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1 is characterized in that: The microseismic event spatiotemporal distribution characteristic diagram in step 6 includes a distribution characteristic diagram of the microseismic event along the tunnel axis and a distribution characteristic diagram of the microseismic event along the tunnel cross section. The specific drawing process is as follows: The large model intelligent agent automatically screens the real microseismic events in the microseismic event data, and uses the X-axis, Y-axis, Z-axis coordinates and energy ratio of the selected microseismic events to draw a distribution characteristic diagram of the microseismic events along the tunnel axis, which is used to represent the density and distribution law of microseismic events at different positions along the tunnel axis. Specifically, spheres of different colors are used to represent events at different time points, where the volume of the sphere is proportional to the energy ratio. The larger the energy ratio, the larger the volume of the sphere. Draw a distribution characteristic map of microseismic events along the tunnel cross section; used to show the density and distribution law of microseismic events in the tunnel lateral direction; The specific drawing process of the microseismic event evolution law diagram is as follows: counting the time and energy ratio of the real microseismic events, calculating the cumulative number of microseismic events and the cumulative microseismic release energy; and using a double-axis line graph to represent it.

6. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1, characterized in that: The step 7 is specifically as follows: the large model intelligent agent disassembles the microseismic event evolution law graph into two single line graphs, namely, the cumulative microseismic event number line graph and the cumulative microseismic release energy line graph, and analyzes the cumulative number of microseismic events at all times and the cumulative microseismic release energy at all times respectively; Calculate and obtain the cumulative number of microseismic events in the warning area and the cumulative energy released by these microseismic events.

7. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1, characterized in that: The rule knowledge base described in step 8 includes feature descriptions and identification rules for different rockburst types; based on these rules, the microseismic event features in the image description are analyzed, the closest rockburst type is determined, and the determination result is output.

8. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1, characterized in that: The step 9 is specifically as follows: if no rockburst occurs between the last daily report and the current daily report, there is no need to upload the on-site rockburst picture, and step 10 is performed directly; otherwise, the on-site rockburst picture that occurs between the last daily report and the current daily report is uploaded; specifically, the name of the on-site rockburst picture is first obtained, and the name is specifically "warning area, rockburst occurrence location, rockburst level, rockburst depth", and then the warning time and area are obtained, that is, "the recording time of the last daily report, the warning area of ​​the last daily report", the excavation time, that is, "the current position of the last daily report~the current position of the current daily report, the time of the last daily report~the time of this daily report", the warning level, that is, "the rockburst warning level of the last daily report", the actual occurrence, that is, "the name of the picture", the construction situation, that is, "the current position of the current daily report~the current position of the last daily report" and the comparative analysis, that is, "whether the actual occurrence is consistent with the warning level".

9. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1, characterized in that: The daily report in step 10 specifically includes a first table, the header of the first table is the warning area information, and the specific location information of the sensor provided by the microseismic monitoring parameter file is also stored in the first table; The first table also includes a microseismic event spatiotemporal distribution characteristic diagram and a microseismic event evolution law diagram, which shows the distribution and characteristics of rock burst and microseismic activities; analyzing the characteristics of microseismic activities, matching and determining the level of rock burst warning according to the picture description and the information in the knowledge base, and storing it in the first table; If no rockburst occurred during the previous daily report, the daily report only contains the first table; otherwise, the second table is added; The second table is used to record and analyze information on rock burst events; the header of the second table uses the pile number segment of the first table recorded in the previous daily report. At the same time, the time recorded in the previous daily report will also be added to the second table as a reference for the warning time and area of ​​this rock burst; the excavation time will be recorded based on the pile number from the last record to the pile number recorded this time, as well as the corresponding time. The warning level will be determined based on the rock burst category judgment result in the previous daily report. The actual rock burst event will be recorded through the name of the picture uploaded by the user, and the construction situation will be reflected by the difference in pile numbers from the previous daily report to the current daily report. The comparative analysis will compare the actual rock burst event with the warning result in the previous daily report to determine the accuracy of the warning and the consistency of the actual occurrence; the second table also contains on-site rock burst pictures uploaded by the user.

10. The method for generating a rockburst microseismic monitoring and early warning report based on a large model agent according to claim 1, characterized in that: The rockburst microseismic monitoring and early warning method based on the large model intelligent agent is implemented based on the rock mechanics knowledge interaction system and specifically includes: a data acquisition module, a data processing module, a rule knowledge base module and a daily report generation module; The data processing module is used to receive the daily report template, TBM parameter file, real-time data of microseismic monitoring sensors, and microseismic event data, obtain the current pile number, warning area, sensor location, and transmit them to the data processing module; The data processing module draws the spatiotemporal distribution characteristic diagram of microseismic events and the evolution law diagram of microseismic events through the large model intelligent body, and performs image analysis and summary through the multimodal large model, and at the same time calculates the cumulative number of microseismic events generated in the warning area and the cumulative energy released by these microseismic events; the rule knowledge base module determines the closest rock burst type through the rule knowledge base module, and outputs the judgment result to the daily report generation module; the daily report generation module generates daily reports through the large model intelligent body.

Images