Coal mine underground rock burst danger assessment system and method based on AI analysis

Abstract

The application discloses a coal mine underground rock burst danger evaluation system and method based on AI analysis in the technical field of coal mine safety monitoring and intelligent risk evaluation, and comprises the following steps: obtaining an initial danger level of a target evaluation area and microseismic monitoring data of a corresponding mine area of the target evaluation area within a set monitoring period; analyzing the microseismic monitoring data to identify all energy level jump type microseismic transition modes appearing in the target evaluation area; determining the evolution time length of each energy level jump type microseismic transition mode and counting the proportion of abnormal modes with an evolution time length less than an evolution time threshold. The application identifies energy level jump type microseismic transition modes meeting energy and time double threshold conditions, and constructs a dynamic correction mechanism for rock burst risk levels based on the relative deviation of the evolution time length of abnormal modes relative to the reference time length of a reference area, so that the phased change of the regional risk level can be effectively responded.

Description

Technical Field

[0001] This invention relates to an AI-based assessment system and method for assessing the risk of rockburst in underground coal mines, belonging to the field of coal mine safety monitoring and intelligent risk assessment technology. Background Technology

[0002] In coal mine production, rockbursts, as a typical type of dynamic disaster, are characterized by their suddenness, destructiveness, and difficulty in early warning, seriously threatening the lives of mine workers and the continuity of equipment operation. To effectively predict rockburst risks, existing technologies typically rely on microseismic monitoring systems to collect real-time rock mass disturbance signals within the mine, and then classify risk levels and provide early warning interventions based on static characteristic parameters such as source energy, magnitude distribution, and event frequency. Some methods also classify and evaluate microseismic data by constructing empirical rules or expert scoring models, thereby forming preliminary hazard level determination results.

[0003] However, existing risk assessment methods have significant shortcomings in addressing the dynamic evolution of mine pressure risks. On the one hand, most methods only focus on static parameter indicators of individual events, lacking the ability to structurally model the evolutionary trends between event sequences, and failing to reveal the abrupt process of microseismic events from low-energy accumulation to high-energy release. On the other hand, the initial risk level often relies on statistical characteristics or rule templates within a single period, lacking a basis for dynamic correction, making the assessment results susceptible to isolated anomaly disturbances and unable to effectively respond to the phased changes in regional risk levels. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide an assessment system and method for rockburst risk in coal mines based on AI analysis. By identifying energy level leap-type microseismic transition modes that meet the dual threshold conditions of energy and time, and based on the relative deviation of the evolution time of the anomalous mode from the reference time of the reference area, a dynamic correction mechanism for rockburst risk level is constructed, which can effectively respond to the phased changes in regional risk level.

[0005] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0006] In a first aspect, the present invention provides a method for assessing the risk of underground rockburst in coal mines based on AI analysis, including:

[0007] Obtain the initial hazard level of the target assessment area and the microseismic monitoring data of the corresponding mining area within the set monitoring period;

[0008] Analyze microseismic monitoring data to identify all energy level jump-type microseismic transition modes appearing within the target assessment area;

[0009] The evolution duration of each energy level transition microseismic transition mode is determined, and the proportion of anomalous modes with an evolution duration less than the evolution time threshold is counted. When the proportion exceeds the preset ratio, a reference area with consistent properties and in a stable state with the target assessment area is selected based on the microseismic monitoring data.

[0010] Obtain the energy level jump-type microseismic evolution mode corresponding to the reference region, and calculate the average value of its evolution duration as the reference duration;

[0011] Based on the relative deviation between the evolution duration of each abnormal pattern and the baseline duration, a correction coefficient is generated, and the initial risk level is corrected based on the correction coefficient.

[0012] Furthermore, the target assessment area refers to the specific spatial location underground in a coal mine where the rockburst hazard level needs to be determined. This location is preset by the mine dispatch system or dynamically delineated based on key areas of concern fed back by the rockburst monitoring system. The initial hazard level represents the risk baseline state of the target assessment area at the beginning of the assessment period, obtained through rockburst early warning technology. The set monitoring period is the microseismic monitoring time window of concern during the risk assessment process, set according to the mine operation rhythm, tectonic response cycle, and microseismic event accumulation characteristics. The microseismic monitoring data includes real-time acquisition of microseismic wave signals caused by underground rock mass fracturing and structural disturbances through the microseismic monitoring system deployed in the mine.

[0013] Furthermore, the energy level jump-type microseismic transition mode includes: within a continuous monitoring period, several low-energy microseismic events with energy values ​​below a first threshold occur first, followed by high-energy microseismic events with energy values ​​above a second threshold, and the time interval between the occurrence time of the first low-energy microseismic event and the occurrence time of the first high-energy microseismic event is less than the transition time threshold, and the time interval is defined as the evolution duration.

[0014] Furthermore, the first threshold is determined based on the energy distribution statistics of the long-term risk-free phase in the historical microseismic data of the target mining area; the second threshold is set with reference to the energy release value in historical cases of mine pressure induction or the alarm threshold of the on-site microseismic system; the transition time threshold is the maximum allowable time interval between low-energy microseismic events and high-energy microseismic events.

[0015] Furthermore, the consistent attributes with the target assessment area include the reference area being consistent with the target assessment area in terms of geological structure type, mining technology, support parameters, and monitoring deployment conditions; the stable state includes the reference area not experiencing any rockburst events, not experiencing any abnormal microseismic events with energy exceeding the preset energy level threshold, not implementing any artificial intervention measures related to rockburst risk control, and its historical assessment level consistently not exceeding the preset risk level threshold; the evolution time threshold is determined based on the statistical distribution of the normal pattern evolution duration in historical monitoring data, and can also be empirically corrected by combining typical evolution cycles before rockburst events; the preset ratio is set based on historical mining area statistical data analysis, expert experience, or dynamic model simulation results.

[0016] Furthermore, the energy level jump-type microseismic evolution mode corresponding to the reference region is obtained, and its average evolution duration is calculated as the baseline duration, including:

[0017] Analyze the microseismic monitoring data, identify all energy level jump-type microseismic evolution modes that appear in the reference area within the set monitoring period, and calculate the evolution duration of each identified energy level jump-type microseismic evolution mode;

[0018] The evolution time of the energy level leap type microseismic evolution mode in all reference regions is averaged to obtain the average evolution time, which is then used as the baseline time.

[0019] Furthermore, based on the relative deviation between the evolution duration of each anomalous pattern and the baseline duration, a correction coefficient is generated, and the initial risk level is corrected based on the correction coefficient, including:

[0020] The relative deviation between the evolution duration of each abnormal pattern and the baseline duration is quantified to obtain several deviation amplitudes;

[0021] All deviation values ​​are averaged to obtain the average deviation value, and a correction coefficient is generated based on the average deviation value.

[0022] The initial risk level is adjusted based on the adjustment factor, and the adjustment model used is as follows:

[0023] in, The revised risk level. The initial risk level, This is a preset adjustment coefficient used to control the magnitude of risk level corrections. This represents the total number of abnormal patterns. For the first The evolution time of each abnormal pattern This is the base duration.

[0024] Secondly, this invention provides an AI-based assessment system for the risk of underground rockburst in coal mines, including:

[0025] Data acquisition module: used to obtain the initial hazard level of the target assessment area and to obtain the microseismic monitoring data of the corresponding mining area within the set monitoring period;

[0026] Pattern recognition module: used to analyze microseismic monitoring data and identify all energy level jump-type microseismic transition modes appearing in the target assessment area;

[0027] Reference area determination module: used to determine the evolution time of each energy level leap-type microseismic transition mode, and to count the proportion of anomalous modes whose evolution time is less than the evolution time threshold. When the proportion exceeds the preset ratio, a reference area with consistent attributes and in a stable state with the target assessment area is selected based on the microseismic monitoring data.

[0028] Reference duration calculation module: used to obtain the energy level leap type microseismic evolution mode corresponding to the reference area, and calculate the average value of its evolution duration as the reference duration;

[0029] Level Correction Module: This module generates correction coefficients based on the relative deviation between the evolution duration of each anomalous pattern and the baseline duration, and then corrects the initial risk level based on these correction coefficients.

[0030] Thirdly, the present invention provides an AI-based assessment device for assessing the risk of underground rockburst in coal mines, including a processor and a storage medium;

[0031] The storage medium is used to store instructions;

[0032] The processor is configured to operate according to the instructions to perform the steps of the method according to any of the foregoing.

[0033] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0034] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0035] This invention identifies energy-level jump-type microseismic transition modes that meet both energy and time threshold conditions. Based on the relative deviation of the evolution duration of the anomalous mode relative to the baseline duration of a reference area, a dynamic correction mechanism for rockburst risk levels is constructed. This mechanism identifies anomalous modes and introduces a reference area with consistent properties and in a stable state. It uses the relative deviation to generate correction coefficients to adjust the initial risk level, enabling the assessment results to more accurately reflect the microseismic evolution state of the target assessment area within a set monitoring period. Compared to traditional static assessment methods, this invention introduces temporal evolution information and a reference comparison mechanism in risk identification, improving the targeting and dynamic adaptability of the assessment process. Attached Figure Description

[0036] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0037] Figure 1 This is a flowchart illustrating the AI-based method for assessing the risk of underground rockburst in coal mines, as provided in Embodiment 1 of the present invention.

[0038] Figure 2 A flowchart illustrating the calculation of the baseline time for the AI-based assessment method for assessing the risk of underground rockburst in coal mines, provided in Embodiment 1 of the present invention.

[0039] Figure 3 The flowchart illustrates the correction of the initial risk level in the AI-based assessment method for underground rockburst hazards in coal mines, as provided in Embodiment 1 of the present invention.

[0040] Figure 4 This is an application architecture diagram of the AI-based assessment system for underground rockburst hazards in coal mines provided in Embodiment 2 of the present invention.

[0041] Figure 5 This is a structural block diagram of the benchmark duration calculation module of the AI-based coal mine rockburst hazard assessment system provided in Embodiment 2 of the present invention;

[0042] Figure 6 This is a structural block diagram of the level correction module of the coal mine rockburst hazard assessment system based on AI analysis provided in Embodiment 2 of the present invention. Detailed Implementation

[0043] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0044] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this invention is for describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0045] Example 1:

[0046] Please see Figure 1 This embodiment proposes an AI-based method for assessing the risk of rockburst in coal mines, specifically including the following steps:

[0047] Step S100: Obtain the initial hazard level of the target assessment area and obtain the microseismic monitoring data of the corresponding mining area within the set monitoring period.

[0048] In this embodiment, the target assessment area refers to the specific spatial location underground in a coal mine where the rockburst hazard level needs to be determined. This is typically a key stress concentration area such as the working face, roadway intersections, or deep mining areas. This area can be preset by the mine dispatch system or dynamically delineated based on key areas of concern reported by the rockburst monitoring system for subsequent data collection and risk modeling.

[0049] The initial hazard level represents the baseline risk status of the target assessment area at the start of the assessment period. It can be obtained using existing rockburst early warning technologies, such as those based on key parameters like microseismic event frequency, energy release, magnitude distribution, and stress concentration factor. These parameters can be generated using expert experience, scoring models, fuzzy logic systems, or AI risk classification algorithms, serving as the foundation for assessing the area's current risk level. This initial level can be a qualitative classification (e.g., Level I to Level V) or a continuous risk score.

[0050] This initial hazard level, serving as the input basis for this invention, is closely related to the core theme of "underground rockburst hazard in coal mines," reflecting that the object addressed by this invention is a further intelligent correction and assessment enhancement based on existing risk awareness. In this solution, the initial level is not a replacement, but rather an object to be corrected, adjusted in conjunction with the deviation characteristics of the evolutionary pattern, so that the assessment results are more consistent with the current evolutionary trend and potential shock-induced situation.

[0051] The monitoring period is defined as the microseismic monitoring time window that this invention focuses on during the risk assessment process. It is typically set based on factors such as the mining operation rhythm, tectonic response cycle, and microseismic event accumulation characteristics. It can be a fixed period (e.g., 24 hours, 7 days, 30 days) or a dynamic sliding period (e.g., rolling data within 48 hours). This monitoring period is used to collect a sufficient number of microseismic events to support the identification and statistical analysis of energy level-elevation-type microseismic transition modes.

[0052] Microseismic monitoring data originates from microseismic monitoring systems deployed in the mine (such as distributed acoustic emission sensors and seismic wave sensor arrays). These systems collect microseismic wave signals in real time caused by rock mass fracturing and structural disturbances underground. The data should at least include elements such as event occurrence time, energy value, magnitude, frequency characteristics, source location, and duration. It can be further expanded to include multi-source information such as ground stress distribution, mining progress, and support deformation monitoring data, providing a multi-dimensional input basis for subsequent analysis.

[0053] Step S200: Analyze the microseismic monitoring data and identify all energy level jump-type microseismic transition modes appearing in the target assessment area. An energy level jump-type microseismic transition mode refers to the following: within a continuous monitoring period, several low-energy microseismic events with energy values ​​below the first threshold occur first, followed by high-energy microseismic events with energy values ​​above the second threshold. The time interval between the occurrence of the first low-energy microseismic event and the occurrence of the first high-energy microseismic event is less than the transition time threshold. This time interval is defined as the evolution duration.

[0054] In this embodiment, by analyzing microseismic monitoring data, an energy-level jump-type microseismic transition mode is identified within the target assessment area. The aim is to extract key evolutionary features that characterize the potential transition trend of rockbursts, further identifying anomalous evolution processes and serving as the core basis for risk level correction. This mode reflects the abrupt transition of microseismic activity from a low-intensity, high-frequency background state to a high-energy release state, and is one of the common precursory manifestations of rockbursts.

[0055] This type of microseismic evolution process lacks systematic identification and quantitative definition in existing technologies. Traditional risk assessment methods mostly rely on static characteristics such as energy, frequency, and density of single-point events, making it difficult to accurately capture the temporal correlation between events and their potential energy level transition trends. This invention, for the first time, constructs a class of microseismic transition modes with structural and critical characteristics by setting time and energy thresholds, thus overcoming the problem of insufficient characterization of "state abrupt change processes" in existing studies, and possessing significant innovation and adaptability.

[0056] The first threshold is used to limit the upper limit energy value of low-energy microseismic events. It can be determined based on the energy distribution statistics of the long-term risk-free phase in the historical microseismic data of the target mining area. It is usually set as the upper quartile or the mean plus the deviation range of the microseismic energy distribution in that phase to ensure that the selected events are indeed "low-energy background".

[0057] The second threshold is used to limit the lower energy value of high-energy microseismic events, representing the critical energy level that may induce an impact response. Its setting can refer to the energy release values ​​in historical cases of mine pressure-induced events or the alarm threshold of the on-site microseismic system to ensure that the identified uplift events have structural disturbance effects and catastrophic potential.

[0058] The transition time threshold is used to limit the maximum permissible time interval between low-energy microseismic events and high-energy microseismic events, and is a key criterion for determining whether an energy-level transition mode of microseismic events is valid. If this time interval exceeds the transition time threshold, it indicates that the energy level evolution process is too slow and does not have obvious transition characteristics, and therefore does not constitute a transition mode.

[0059] Step S300: Determine the evolution duration of each energy level transition microseismic transition mode, and count the proportion of anomalous modes whose evolution duration is less than the evolution time threshold. When the proportion exceeds the preset ratio, a reference area with consistent attributes and in a stable state with the target assessment area is selected based on the microseismic monitoring data.

[0060] In the process of selecting reference areas, having the same attributes as the target assessment area means that the reference area corresponds to the target assessment area in terms of geological structure type, mining technology, support parameters and monitoring deployment conditions; being in a stable state means that the reference area has not experienced any rockburst events, has not experienced any abnormal microseismic events with energy exceeding the preset energy level threshold within the set monitoring period, has not implemented any artificial intervention measures related to rockburst risk control, and its historical assessment level has always been no higher than the preset risk level threshold.

[0061] In this embodiment, the evolution time threshold is used to further screen for anomalous energy level jump-type microseismic transition modes with significant acceleration characteristics. Although the jump time threshold is already used as a configurational condition for determining whether such a transition mode exists, the evolution time threshold is a further refinement to identify cases where the evolution process is abnormally rapid. This threshold can be determined based on the statistical distribution of the evolution duration of normal modes in historical monitoring data, for example, by setting it as the lower quartile or an empirical threshold, or by empirically adjusting it in conjunction with the typical evolution cycle before a rockburst event. By limiting this threshold, anomalous modes potentially in a critical jump state can be identified as a focus of attention during the risk correction process.

[0062] The preset percentage is used to determine the statistical significance of abnormal patterns in the target assessment area and serves as a threshold indicator for deciding whether to trigger the correction mechanism. Its setting can be based on historical mining area statistical data analysis, expert experience, or dynamic model simulation results, and is typically between 20% and 40% to balance identification sensitivity and false alarm control. If this percentage is set too low, it may easily lead to misjudgments due to occasional anomalies; if the percentage is too high, it may mask the true overall upward trend of risk.

[0063] When the proportion of anomalous patterns exceeds a preset percentage, it indicates that a significant proportion of microseismic energy level transition events within the current target assessment area exhibit an abnormally accelerated evolution rate, potentially suggesting that the regional rock mass stress state is approaching an unstable boundary. In this case, the initial hazard level is insufficient to accurately reflect the true risk level of the region. Therefore, it is necessary to introduce a reference area and correct the current level through deviation analysis, thereby improving the timeliness and accuracy of the assessment results. Thus, this percentage determination serves as a prerequisite for the correction process, effectively preventing over-response to single, short-term, sudden anomalies and providing a statistical basis for the correction mechanism.

[0064] In the process of selecting reference areas, it is necessary to ensure that they have strong comparability and stability with the target assessment area. Having consistent attributes means maintaining a high degree of consistency with the target assessment area in multiple aspects, such as geological structure type (e.g., fault structure, surrounding rock properties), mining technology (e.g., advance direction, mining height control), support parameters (e.g., anchor bolt type, support density), and monitoring deployment conditions (e.g., sensor density, deployment angle), to ensure that the extracted benchmark duration has structural consistency and engineering applicability.

[0065] The term "stable state" refers to a situation where the reference area experiences no significant shock-inducing factors within the set monitoring period. Specifically, this includes no rockburst events, no micro-seismic disturbances exceeding a set threshold in energy, no implementation of any artificial intervention measures such as pressure relief, reverse adjustment, or advanced support, and the risk level remains within a safe control range (e.g., not exceeding the system's set warning lower limit). This screening mechanism ensures that the collected baseline duration originates from typical and safe natural evolution processes, avoiding the introduction of interfering reference data.

[0066] Step S400: Obtain the energy level jump-type microseismic evolution mode corresponding to the reference region, and calculate its average evolution duration as the baseline duration. Please refer to [link to relevant documentation]. Figure 2 To obtain the energy level jump-type microseismic evolution mode corresponding to the reference region and calculate its average evolution duration as the baseline duration, the specific steps include:

[0067] Step S401: Analyze the microseismic monitoring data, identify all energy level jump-type microseismic evolution modes that appear in the reference area within the set monitoring period, and calculate the evolution duration of each identified energy level jump-type microseismic evolution mode.

[0068] Step S402: Average the evolution time of the energy level jump type microseismic evolution mode in all reference regions to obtain the average evolution time, and use it as the benchmark time.

[0069] In this embodiment, step S401 specifically involves analyzing the microseismic monitoring data of the reference area within a set monitoring period and extracting all event sequences that satisfy the configuration characteristics of an energy level jump-type microseismic evolution mode according to a unified event identification rule. Specifically, the system first filters out all low-energy microseismic events with energy values ​​below a first threshold, and clusters them according to their chronological order to identify consecutive segments. Then, it searches for high-energy microseismic events with energy values ​​above a second threshold within a short time window after each segment. If the time interval between the two types of events is less than the transition time threshold, it is determined to be a valid energy level jump-type microseismic evolution mode. For each identified evolution mode, the time interval between the occurrence time of its first low-energy microseismic event and the occurrence time of its first high-energy microseismic event is recorded as the evolution duration of the mode.

[0070] In step S402, the evolution durations of the identified energy level jump-type microseismic evolution modes in all reference areas are statistically processed. The average evolution duration is calculated using an arithmetic mean, and this value is used as the baseline duration. The average value is chosen as the baseline because the reference areas are in a stable state within the set monitoring period, and their microseismic evolution process represents typical evolutionary characteristics under non-abnormal conditions, possessing statistical representativeness and reliability. Using this average value as the baseline helps to measure the relative deviation between each anomalous mode and the normal evolution level in the target assessment area. This baseline duration reflects the normal level of the natural stress release rhythm in the area and provides a unified reference for subsequent deviation amplitude calculation and correction factor construction, ensuring the objectivity and data consistency of the risk level correction process.

[0071] Step S500: Based on the relative deviation between the evolution duration of each abnormal pattern and the baseline duration, generate a correction coefficient, and adjust the initial risk level based on the correction coefficient. Please refer to [link / reference]. Figure 3 Based on the relative deviation between the evolution duration of each abnormal pattern and the baseline duration, a correction coefficient is generated, and the initial risk level is corrected based on the correction coefficient. Specifically, the steps include:

[0072] Step S501: Quantify the relative deviation between the evolution duration of each abnormal mode and the baseline duration to obtain several deviation amplitudes;

[0073] Step S502: Average all deviation values ​​to obtain the average deviation value, and generate a correction coefficient based on the average deviation value;

[0074] Step S503: Adjust the initial risk level according to the correction factor.

[0075] When revising the initial risk level, a preset revision model is used;

[0076] The corrected model is as follows: ;

[0077] in This refers to the revised risk level. This refers to the initial risk level. This refers to the total number of abnormal patterns. It refers to the first The evolution time of each abnormal pattern This refers to the base duration. It refers to the first The deviation magnitude of each abnormal pattern relative to the baseline duration. This refers to the correction factor, i.e., the average deviation amplitude. This refers to the preset adjustment coefficient used to control the magnitude of risk level corrections, and Greater than zero.

[0078] In this embodiment, the relative deviation between the evolution duration of each anomalous mode and the baseline duration is quantified, and the average value of all deviation amplitudes is calculated. A correction coefficient is then generated based on this average deviation amplitude to correct the initial risk level, which has clear physical meaning and computational rationality. This process not only preserves the quantitative perception capability of microseismic evolution trends but also eliminates the interference of individual anomalous fluctuations through statistical convergence, making the correction results more robust and representative.

[0079] The core rationale behind this method is that when abnormal patterns generally exhibit evolution durations significantly shorter than the baseline duration, it indicates that the current target assessment area is undergoing a more intense and rapid energy accumulation and release transition. The original risk level can no longer accurately reflect the actual state of the area, necessitating dynamic adjustment. Constructing a correction coefficient using the average deviation amplitude not only quantitatively characterizes the risk change trend but also allows for flexible control of the correction magnitude through the adjustment coefficient K, balancing sensitivity and stability and avoiding misjudgments caused by a single extreme value.

[0080] The risk correction model formula used in this invention is highly intuitive and practical for engineering applications. By representing the risk level as the result of correcting the product of the initial level and the deviation, the system can dynamically adapt to the current microseismic evolution level in a linear and controllable manner. Compared to black-box neural network models or expert scoring methods, this method has advantages such as transparency and interpretability, adjustable parameters, and reusable structure, making it particularly suitable for mining pressure scenarios with high safety requirements and clear response decisions.

[0081] Furthermore, based on this deviation amplitude, a multi-dimensional weight correction mechanism can be introduced. For example, risk stability can be adjusted according to the distribution pattern of the deviation amplitude (such as variance and skewness), or incremental time weights can be set for the deviation amplitudes at different time periods to construct a time-series weighted risk correction model, thereby achieving stronger perception and finer-grained control over the evolution of rockburst trends. Through these extensions, the risk level correction model of this invention has good adaptability and scalability, and can be widely applied to various downhole microseismic dynamic assessment systems.

[0082] Example 2:

[0083] Please see Figure 4 This embodiment discloses an AI-based assessment system for underground rockburst hazards in coal mines, which can implement the AI-based assessment method for underground rockburst hazards in coal mines described in Embodiment 1. The system includes: a data acquisition module 100, a pattern recognition module 200, a reference area determination module 300, a baseline duration calculation module 400, and a level correction module 500, wherein:

[0084] The data acquisition module 100 is used to obtain the initial hazard level of the target assessment area and to obtain the microseismic monitoring data of the corresponding mining area within the set monitoring period.

[0085] The pattern recognition module 200 is used to analyze microseismic monitoring data and identify all energy level jump-type microseismic transition modes appearing in the target assessment area. An energy level jump-type microseismic transition mode refers to the following: within a continuous monitoring period, several low-energy microseismic events with energy values ​​below a first threshold occur first, followed by high-energy microseismic events with energy values ​​above a second threshold. The time interval between the occurrence of the first low-energy microseismic event and the occurrence of the first high-energy microseismic event is less than the transition time threshold, where this time interval is defined as the evolution duration.

[0086] The reference area determination module 300 is used to determine the evolution duration of each energy level transition microseismic transition mode and to count the proportion of anomalous modes whose evolution duration is less than the evolution time threshold. When the proportion exceeds the preset ratio, a reference area with consistent attributes and in a stable state with the target assessment area is selected based on the microseismic monitoring data.

[0087] In the process of selecting reference areas, having the same attributes as the target assessment area means that the reference area corresponds to the target assessment area in terms of geological structure type, mining technology, support parameters and monitoring deployment conditions; being in a stable state means that the reference area has not experienced any rockburst events, has not experienced any abnormal microseismic events with energy exceeding the preset energy level threshold within the set monitoring period, has not implemented any artificial intervention measures related to rockburst risk control, and its historical assessment level has always been no higher than the preset risk level threshold.

[0088] The reference duration calculation module 400 is used to obtain the energy level jump-type microseismic evolution mode corresponding to the reference region and calculate the average value of its evolution duration as the reference duration. Please refer to [link / reference]. Figure 5 The baseline duration calculation module 400 specifically includes:

[0089] The evolution duration calculation unit 401 is used to analyze microseismic monitoring data, identify all energy level jump-type microseismic evolution modes that appear in the reference area within a set monitoring period, and calculate the evolution duration of each identified energy level jump-type microseismic evolution mode.

[0090] The average duration calculation unit 402 is used to average the evolution duration of the energy level leap type microseismic evolution mode in all reference areas to obtain the average evolution duration, and use it as the reference duration.

[0091] The risk level correction module 500 generates correction coefficients based on the relative deviation between the evolution duration of each anomalous pattern and the baseline duration, and corrects the initial risk level based on these correction coefficients. Please refer to [link / reference]. Figure 6 The level correction module 500 specifically includes:

[0092] The deviation amplitude calculation unit 501 is used to quantify the relative deviation between the evolution duration of each abnormal mode and the reference duration, and obtain several deviation amplitudes;

[0093] The correction coefficient determination unit 502 is used to average all deviation values ​​to obtain the average deviation value, and generate correction coefficients based on the average deviation value.

[0094] The correction factor application unit 503 is used to correct the initial risk level based on the correction factor.

[0095] When revising the initial risk level, a preset revision model is used;

[0096] The corrected model is as follows: ;

[0097] in This refers to the revised risk level. This refers to the initial risk level. This refers to the total number of abnormal patterns. It refers to the first The evolution time of each abnormal pattern This refers to the base duration. It refers to the first The deviation magnitude of each abnormal pattern relative to the baseline duration. This refers to the correction factor, i.e., the average deviation amplitude. This refers to the preset adjustment coefficient used to control the magnitude of risk level corrections, and Greater than zero.

[0098] Example 3:

[0099] This invention also provides an AI-based assessment device for underground rockburst hazards in coal mines, which can realize the AI-based assessment method for underground rockburst hazards in coal mines described in Embodiment 1, including a processor and a storage medium;

[0100] The storage medium is used to store instructions;

[0101] The processor is configured to operate according to the instructions to perform the steps of the following method:

[0102] Obtain the initial hazard level of the target assessment area and the microseismic monitoring data of the corresponding mining area within the set monitoring period;

[0103] Analyze microseismic monitoring data to identify all energy level jump-type microseismic transition modes appearing within the target assessment area;

[0104] The evolution duration of each energy level transition microseismic transition mode is determined, and the proportion of anomalous modes with an evolution duration less than the evolution time threshold is counted. When the proportion exceeds the preset ratio, a reference area with consistent properties and in a stable state with the target assessment area is selected based on the microseismic monitoring data.

[0105] Obtain the energy level jump-type microseismic evolution mode corresponding to the reference region, and calculate the average value of its evolution duration as the reference duration;

[0106] Based on the relative deviation between the evolution duration of each abnormal pattern and the baseline duration, a correction coefficient is generated, and the initial risk level is corrected based on the correction coefficient.

[0107] Example 4:

[0108] This invention also provides a computer-readable storage medium that implements the AI-based assessment method for assessing the risk of underground rockburst in coal mines as described in Embodiment 1. The medium stores a computer program that, when executed by a processor, performs the steps of the following method:

[0109] Obtain the initial hazard level of the target assessment area and the microseismic monitoring data of the corresponding mining area within the set monitoring period;

[0110] Analyze microseismic monitoring data to identify all energy level jump-type microseismic transition modes appearing within the target assessment area;

[0111] The evolution duration of each energy level transition microseismic transition mode is determined, and the proportion of anomalous modes with an evolution duration less than the evolution time threshold is counted. When the proportion exceeds the preset ratio, a reference area with consistent properties and in a stable state with the target assessment area is selected based on the microseismic monitoring data.

[0112] Obtain the energy level jump-type microseismic evolution mode corresponding to the reference region, and calculate the average value of its evolution duration as the reference duration;

[0113] Based on the relative deviation between the evolution duration of each abnormal pattern and the baseline duration, a correction coefficient is generated, and the initial risk level is corrected based on the correction coefficient.

[0114] As is known from common technical knowledge, this invention can be implemented through other embodiments that do not depart from its spirit or essential characteristics. Therefore, the disclosed embodiments described above are merely illustrative and not exhaustive. All modifications within the scope of this invention or its equivalents are included in this invention.

[0115] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0116] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0117] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0118] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An evaluation method of rock burst danger in a coal mine based on AI analysis, characterized in that, include: The initial hazard level of the target assessment area and the microseismic monitoring data of the corresponding mining area within a set monitoring period are obtained. The initial hazard level represents the risk baseline state of the target assessment area at the beginning of the assessment period and is obtained through rockburst early warning technology. The microseismic monitoring data includes real-time acquisition of microseismic wave signals caused by rock mass fracture and structural disturbance in the mine through a microseismic monitoring system deployed in the mine. Analyze microseismic monitoring data to identify all energy level jump-type microseismic transition modes appearing in the target assessment area. The energy level jump-type microseismic transition modes include: within a continuous monitoring period, several low-energy microseismic events with energy values ​​below a first threshold occur first, followed by high-energy microseismic events with energy values ​​above a second threshold. The time interval between the occurrence time of the first low-energy microseismic event and the occurrence time of the first high-energy microseismic event is less than the transition time threshold. The time interval is defined as the evolution duration. The evolution duration of each energy level transition microseismic transition mode is determined, and the proportion of anomalous modes with an evolution duration less than the evolution time threshold is counted. When the proportion exceeds a preset ratio, a reference area with consistent attributes and in a stable state is selected based on microseismic monitoring data. The consistent attributes with the target assessment area include that the reference area corresponds to the target assessment area in terms of geological structure type, mining technology, support parameters, and monitoring deployment conditions. The stable state includes that the reference area has not experienced any rockburst events, has not experienced any anomalous microseismic events with energy exceeding the preset energy level threshold, has not implemented any artificial intervention measures related to rockburst risk control, and its historical assessment level has always been no higher than the preset risk level threshold. Obtain the energy level jump-type microseismic evolution mode corresponding to the reference region, and calculate the average value of its evolution duration as the reference duration; Based on the relative deviation between the evolution duration of each anomalous pattern and the baseline duration, a correction coefficient is generated, and the initial risk level is adjusted based on the correction coefficient, including: The relative deviation between the evolution duration of each abnormal pattern and the baseline duration is quantified to obtain several deviation amplitudes; All deviation values ​​are averaged to obtain the average deviation value, and a correction coefficient is generated based on the average deviation value. The initial risk level is adjusted based on the adjustment factor, and the adjustment model used is as follows: ； in, The revised risk level. The initial risk level, This is a preset adjustment coefficient used to control the magnitude of risk level corrections. This represents the total number of abnormal modes. For the first The evolution time of each abnormal pattern Based on the base duration, This refers to the correction factor.

2. The method for assessing the risk of underground rockburst in coal mines based on AI analysis according to claim 1, characterized in that, The target assessment area refers to the specific spatial location underground in a coal mine where the risk level of rockburst needs to be determined. It is preset by the mine dispatch system or dynamically delineated based on the key areas of concern fed back by the rockburst monitoring system. The set monitoring period is the microseismic monitoring time window that is of concern in the risk assessment process. It is set according to the mining operation rhythm, tectonic response cycle, and microseismic event accumulation characteristics.

3. The method for assessing the risk of underground rockburst in coal mines based on AI analysis according to claim 1, characterized in that, The first threshold is determined based on the energy distribution statistics of the long-term risk-free phase in the historical microseismic data of the target mining area; the second threshold is set with reference to the energy release value in historical cases of mine pressure induction or the alarm threshold of the on-site microseismic system; the transition time threshold is the maximum allowable time interval between low-energy microseismic events and high-energy microseismic events.

4. The method for assessing the risk of underground rockburst in coal mines based on AI analysis according to claim 1, characterized in that, The evolution time threshold is determined based on the statistical distribution of the evolution time of the normal pattern in historical monitoring data, and is empirically corrected by combining the typical evolution cycle before the occurrence of rockburst events; the preset ratio is set based on the analysis of historical mining area statistical data, expert experience, or dynamic model simulation results.

5. The method for assessing the risk of underground rockburst in coal mines based on AI analysis according to claim 1, characterized in that, Obtain the energy level jump-type microseismic evolution mode corresponding to the reference region, and calculate its average evolution duration as the baseline duration, including: Analyze microseismic monitoring data, identify all energy level jump-type microseismic evolution modes that appear in the reference area within a set monitoring period, and calculate the evolution duration of each identified energy level jump-type microseismic evolution mode; The evolution time of the energy level leap type microseismic evolution mode in all reference regions is averaged to obtain the average evolution time, which is then used as the baseline time.

6. An AI-based assessment system for underground rockburst hazards in coal mines, characterized by: A method for assessing the risk of underground rockburst in coal mines based on AI analysis as described in any one of claims 1 to 5, comprising: Data acquisition module: used to obtain the initial hazard level of the target assessment area and to obtain the microseismic monitoring data of the corresponding mining area within the set monitoring period; Pattern recognition module: used to analyze microseismic monitoring data and identify all energy level jump-type microseismic transition modes appearing in the target assessment area; Reference area determination module: used to determine the evolution time of each energy level leap-type microseismic transition mode, and to count the proportion of anomalous modes whose evolution time is less than the evolution time threshold. When the proportion exceeds the preset ratio, a reference area with consistent attributes and in a stable state with the target assessment area is selected based on the microseismic monitoring data. Reference duration calculation module: used to obtain the energy level leap type microseismic evolution mode corresponding to the reference area, and calculate the average value of its evolution duration as the reference duration; Level Correction Module: This module generates correction coefficients based on the relative deviation between the evolution duration of each anomalous pattern and the baseline duration, and then corrects the initial risk level based on these correction coefficients.

7. An AI-based assessment device for assessing the risk of rockburst in coal mines, characterized in that: Including processor and storage media; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1 to 5.

Images