An Adaptive Localization Method for Microseismic Events in Deep-Buried Tunnels

Abstract

This invention provides an adaptive localization method for microseismic events in deeply buried tunnels. It includes an optimization initialization mechanism based on target gradients to accurately locate the initial point. Building upon existing local algorithms, it explores and analyzes the relationship between the seismic source location and the distribution of the sensor array in real engineering environments, as well as the spatial morphology of the objective function value. This summarizes the patterns of the initial point and the objective function gradient, determines the initialization region, optimizes the initial point selection, and maintains the algorithm's convergence rate. A data evaluation strategy is used to process low-quality data affected by the engineering environment. Based on the actual environment of deeply buried tunnel engineering and the two-wave theorem, data filtering and weighting rules are set to ensure that the strategy conforms to the actual engineering background, providing a sufficient and reliable theoretical basis for data elimination and weighting strategies. This invention solves the problems of difficult and complex microseismic event localization in tunnels, achieving accurate localization of microseismic sources and ensuring safe production in deeply buried tunnel engineering.

Description

Technical Field

[0001] This invention relates to the field of microseismic monitoring technology for deep-buried tunnel engineering, and more particularly to an adaptive localization method for microseismic events in deep-buried tunnels. Background Technology

[0002] In recent years, my country's overall national strength has further increased, and its construction has moved towards more comprehensive and in-depth development. Simultaneously, the demand for underground energy is growing, leading to a surge in deep engineering projects. However, under high ground stress or deep-buried tunnels, the disturbance caused by excavation at the working face often leads to a redistribution of rock stress, rock fracturing, and geological disasters such as collapses, rockbursts, and water inrushes. These disasters seriously threaten the safety of construction personnel and equipment, and affect construction progress. Therefore, effectively monitoring and accurately locating the microseismic source during rockbursts, micro-earthquakes, or strong earthquakes is of great significance for the prediction and effective protection against rockbursts.

[0003] Rockburst prediction is a crucial aspect of earthquake prevention and disaster reduction, providing strong guidance for deep engineering work sites. Through extensive research and engineering practice, microseismic monitoring has become an important tool. Sensors embedded within the rock mass capture and analyze vibration signals generated during rock fracturing to obtain temporal, spatial, and intensity information such as the time, location, and energy of the fracturing event. Therefore, effectively utilizing microseismic monitoring technology to monitor the surrounding rock in real time, accurately locate microseismic sources, and analyze the effective characteristics of the captured vibration signals is key to quantitatively determining whether microfractures are hazardable.

[0004] Currently, there is no effective method for locating the microseismic sources that frequently occur during the mining of deep-buried tunnels. The specific challenges are as follows:

[0005] (1) Difficulty in Location: For monitoring deeply buried tunnels, extensive practical experience has shown that the locations of microseismic sources are usually distributed outside the sensor array, making it difficult to capture and accurately locate the sources. Many experts have proposed solutions to address this difficulty, but they all have limitations. For example, the proposed PSO swarm intelligence algorithm can improve the accuracy of source location outside the array, but its search time is slow and it is prone to getting trapped in local optima. Another widely used algorithm is the local linear optimization algorithm, which improves the accuracy of source location prediction by continuously optimizing the initial point selection, but the initial point is unknown. Therefore, there is an urgent need to provide an adaptive location method for microseismic events in deeply buried tunnels that can be continuously optimized based on existing algorithms to achieve precise source location.

[0006] (2) Data Complexity: Due to the complex environment of deep-buried tunnel engineering, such as adverse geological conditions, diffraction from voids, and noise, the data obtained by sensors is currently extremely complex. The noise from the complex mining environment and the errors caused by complex rock types often introduce certain errors or even erroneous values ​​into the algorithm. Therefore, how to denoise the data obtained by the sensors to obtain a "cleaner" data signal is an important aspect of improving prediction performance. Summary of the Invention

[0007] To address the aforementioned technical challenges of locating microseismic events in tunnels, including difficulties and complex data collection, this invention provides an adaptive localization method for microseismic events in deeply buried tunnels. Building upon existing local algorithms, this method explores and analyzes the relationship between the seismic source location and the distribution of the sensor array in a real engineering environment, as well as the spatial morphology of the objective function value. It summarizes the patterns of the initial point and the gradient of the objective function, and applies an optimization initialization mechanism based on the objective gradient to optimize the localization of the initial point. This not only maintains the algorithm's convergence rate but also improves its localization accuracy. Furthermore, considering the characteristics of sensor arrangement in deeply buried tunnel engineering, different data rejection thresholds are proposed for sensors in the same row and different rows. A voting algorithm is used to eliminate low-quality data within the engineering context.

[0008] The technical means employed in this invention are as follows:

[0009] An adaptive localization method for microseismic events in deeply buried tunnels includes:

[0010] The optimization initialization mechanism based on the target gradient is used to accurately locate the initial point;

[0011] By using a data assessment strategy, low-quality data affected by the engineering environment is processed to identify dangerous areas for excavation activities and accurately predict rockburst locations.

[0012] Furthermore, the target gradient-based optimization initialization mechanism for accurate initial point localization includes:

[0013] Analyze the distribution pattern of the initial points;

[0014] Analyze the spatial distribution of the objective function;

[0015] Compare the optimization trajectories with different initial points;

[0016] Investigate the relationship between the initialization region and the gradient;

[0017] Determine the initialization region;

[0018] Determine the initial point.

[0019] Furthermore, the analysis of the distribution patterns of the initial points includes: classifying the earthquake source distribution areas according to the relevant positions of the sensor array and the earthquake source, mainly into three categories: earthquake source on the left side of the array, earthquake source in the middle of the array, and earthquake source on the right side of the array; and analyzing the distribution patterns of the initial points under different earthquake source distribution areas.

[0020] When the seismic source is on the left side of the array, the initialization regions are all concentrated on the left side of the array.

[0021] When the seismic source is in the middle of the array, the initialization area is concentrated in the middle of the array;

[0022] When the seismic source is on the right side of the array, the initialization region is concentrated in the right side region of the array.

[0023] Furthermore, the analysis of the spatial distribution of the objective function includes: based on the similarity between the spatial value distribution of the three-dimensional objective function on the z-axis sectional plane and the spatial distribution of the two-dimensional objective function, extrapolating the two-dimensional patterns to the three-dimensional application; and analyzing the relationship between the spatial distribution shape of the two-dimensional objective function values ​​and the initial point distribution.

[0024] When the actual microseismic event is located in the left region of the array, the objective function value shows a characteristic of being smaller on the left and larger on the right, and the objective function value on both sides changes little, while the objective function value in the middle of the array changes significantly.

[0025] When a real microseismic event is located in the middle of the array, the objective function value exhibits a characteristic of being larger on both sides and smaller on the right side and smaller in the middle. Furthermore, the objective function value on both sides changes relatively little, while the objective function value in the middle of the array changes significantly.

[0026] When a real microseismic event is located in the right region of the array, the objective function value exhibits a characteristic of being larger on the left and smaller on the right, with the objective function values ​​on both sides changing little, while the objective function value in the middle of the array changes significantly.

[0027] Furthermore, the comparison of the optimization trajectories at different initial points includes:

[0028] Different initial points were selected for iterative optimization. By comparing the optimization trajectories of different initial points, it was found that the positioning results of different initial points were different. Furthermore, the gradient of the objective function of the initial point that could be correctly positioned was large, while the gradient of the objective function of the initial point that could not be correctly positioned was small. Therefore, further optimization of the selection of initial points can be achieved by looking for patterns in the gradients.

[0029] Furthermore, the study of the relationship between the initialization region and the gradient includes: based on the contour map of the earthquake source distribution, the gradient corresponding to the objective function, and the function value, summarizing the patterns to obtain the relationship between the initialization region and the gradient.

[0030] If the negative gradient direction points to the left, the initialization region is located in the left region of the sensor array;

[0031] If the negative gradient direction points to the middle, then the initialization region is located in the middle of the sensor array;

[0032] If the negative gradient direction points to the right, then the initialization region is located in the right region of the sensor array.

[0033] Furthermore, determining the initialization region includes: calculating the negative gradient direction at the midpoint of each row of sensors, and taking the side that all gradients point to as the initialization region; the rules for determining the initialization region are as follows:

[0034] Let the angles between the negative gradient direction at the midpoint of the two rows of sensors and the x-axis be α1 and α2, respectively. Then we have:

[0035] Satisfaction The initialization region is located in the left region of the sensor array;

[0036] Satisfaction The initialization area is located in the middle of the sensor array;

[0037] Satisfaction The initialization region is located in the right region of the sensor array.

[0038] Further, determining the initial point includes: initializing the target point within the initialization region, and obtaining the initial point by weighting the coordinates of the four vertices of the region based on the principle that the gradient value is smaller the closer it is to the minimum value; the formula for calculating the initial point is:

[0039]

[0040] In the formula, Represents weight, d i represents the gradient magnitude, and n represents the four vertices of the region.

[0041] Furthermore, the process of using a time-based data evaluation strategy to process low-quality data affected by the engineering environment, identify hazardous areas for excavation activities, and accurately predict rockburst locations includes:

[0042] Determine the threshold based on the maximum possible distance;

[0043] Data is removed based on the maximum threshold.

[0044] Based on the two-wave theorem, adaptive weighting;

[0045] The location of microseismic events is obtained by using an adaptive weighted objective function and an optimization solution method.

[0046] Furthermore, determining the threshold based on the maximum possible distance includes: proposing different data rejection thresholds for sensors in the same row and different rows, based on the characteristics of sensor arrangement in deep-buried tunnel engineering.

[0047] Assume the distance between two rows of sensors is 'a', and the distances between adjacent sensors in the same row are 'b' and 'c'.

[0048] Different rows of sensors: For any seismic source, the time difference Δt between any two P-waves received by different rows of sensors. ij All less than T different_max ,Right now In the formula, S i S j ′ represents the distance between any two sensors in different rows;

[0049] For any seismic source, the time difference Δt between any two P-waves received by any two sensors in the same row of sensors is the same. ij All less than T same_max ,Right now In the formula, S i S j This represents the distance between any two sensors in the same row.

[0050] Furthermore, the step of removing data based on a maximum threshold includes: deriving the maximum threshold and then using a voting algorithm to remove data. The specific operation process is as follows:

[0051] The sensor data is divided into the same row and different rows. Any two sensors are selected in turn and their positions are determined to determine the maximum threshold. The arrival time difference between the two sensors is calculated.

[0052] If the time difference is less than the maximum threshold, the credit value is incremented by 1, and this step is repeated until all sensor combinations are traversed.

[0053] All sensors are ranked according to their credit scores, and the bottom 20% of sensors are removed. The remaining sensors are retained for further processing.

[0054] Furthermore, the adaptive weighting based on the two-wavelength theorem includes: obtaining the distance using the time difference between P and S based on the two-wavelength theorem, and using the distance to assist in data quality evaluation. Specifically, the operation is as follows:

[0055] The objective function is improved to Using this weighted adaptive objective function, the microseismic event location results are obtained through an optimized solution method; where, As the basis for adaptive weighting of data, it is obtained based on the spatial relationship theorem. Therefore, the data filtering rules are set as follows: The data is unreliable, and The larger the ratio, the less reliable the data; the formula for calculating Δd is: Δd=|d1-d2|, where d1 and d2 represent the distances between the two sensors and the microseismic event, ΔD xyzThis represents the third edge in the spatial relationship between the sensor and the microseismic event.

[0056] Compared with the prior art, the present invention has the following advantages:

[0057] 1. The adaptive localization method for microseismic events in deep-buried tunnels provided by this invention overcomes the shortcomings of existing algorithms in monitoring and accurately locating microseismic events in tunnels, and achieves precise localization of microseismic sources, effectively ensuring the safe production of deep-buried tunnel projects.

[0058] 2. The adaptive localization method for microseismic events in deep-buried tunnels provided by this invention, based on existing local algorithms, explores and analyzes the relationship between the source location and the distribution of the sensor array in the real engineering environment and the spatial shape of the objective function value, summarizes the rules of the initial point and the gradient of the objective function, determines the initialization area, and applies an optimization initialization mechanism based on the objective gradient to locate the initial point well. This not only improves the positioning accuracy of the algorithm, but also maintains the convergence rate of the algorithm.

[0059] 3. The adaptive positioning method for microseismic events in deep-buried tunnels provided by this invention processes "low-quality" data affected by the engineering environment through a time-based data evaluation strategy. Based on the actual environment of deep-buried tunnel engineering and the two-wave theorem, the data screening rules are set, which not only ensures that the strategy conforms to the actual engineering background, but also provides a sufficient and reliable theoretical basis for data removal.

[0060] 4. The adaptive localization method for microseismic events in deep-buried tunnels provided by this invention achieves precise localization of the final seismic source, effectively identifies dangerous areas of excavation activities, and enables accurate prediction of rockburst locations.

[0061] Based on the above reasons, this invention can be widely promoted in fields such as microseismic monitoring of deep-buried tunnel engineering. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0063] Figure 1 This is a flowchart of the method of the present invention.

[0064] Figure 2 A flowchart of a data removal method based on a maximum threshold provided in an embodiment of the present invention.

[0065] Figure 3This is a schematic diagram illustrating the spatial relationship between a pair of sensors and a microseismic event, provided in an embodiment of the present invention. Detailed Implementation

[0066] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0067] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0069] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following figures denote similar items; therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0070] like Figure 1 As shown, this invention provides an adaptive localization method for microseismic events in deeply buried tunnels, comprising:

[0071] S1. An optimization initialization mechanism based on the target gradient is used to locate the initial point well. In this embodiment, the distribution characteristics of the objective function under the influence of the relative position of the seismic source and the array are analyzed in combination with the layout characteristics of the site. The distribution law of the initial point that can be correctly located under different objective functions is revealed, and a good initial point distribution is obtained. A gradient-based initialization point selection mechanism is proposed to complete the good location of the initial point.

[0072] S2. Low-quality data affected by the engineering environment is processed using a data evaluation strategy to identify hazardous areas for excavation activities and accurately predict rockburst locations. In this embodiment, data filtering rules are set based on the dual-wave theorem, the spatial relationship between sensors and microseismic events, and other practical engineering backgrounds. Considering the deployment characteristics of sensors on-site, different thresholds are used for sensors in the same row and those in different rows, avoiding the use of a single threshold. Furthermore, considering the influence of various on-site factors and air diffraction, and combining the characteristics of the on-site data, the maximum threshold is appropriately adjusted and used as the standard for deleting abnormal data. Finally, a voting algorithm is used to further filter the data. A reliable data processing strategy is used to denoise the data affected by various on-site factors, avoiding prediction errors caused by "dirty" data. The microseismic event location results are obtained by designing an adaptive weighted objective function.

[0073] In a specific implementation, as a preferred embodiment of the present invention, step S1 specifically includes:

[0074] S11. Analyze the distribution pattern of the initial points;

[0075] S12. Analyze the spatial distribution of the objective function;

[0076] S13. Compare the optimization trajectories with different initial points;

[0077] S14. Study the relationship between the initialization region and the gradient;

[0078] S15. Determine the initialization region;

[0079] S16. Determine the initial point.

[0080] In a specific implementation, as a preferred embodiment of the present invention, step S11 specifically includes: classifying the earthquake source distribution area according to the relevant positions of the sensor array and the earthquake source, mainly into three categories: earthquake source on the left side of the array, earthquake source in the middle of the array, and earthquake source on the right side of the array; analyzing the distribution pattern of the initial point under different earthquake source distribution areas:

[0081] S111. When the seismic source is on the left side of the array, the initialization regions are all concentrated in the left side of the array. However, as long as the seismic source is on the left side of the array, there is a common good initialization region in its initialization regions, that is, the initialization regions are all concentrated in the left side of the array.

[0082] S112. When the seismic source is in the middle of the array, the initialization regions are all concentrated in the middle of the array; however, as long as the seismic source is in the middle of the array, there is a common good initialization region in its initialization region, that is, the initialization regions are all concentrated in the middle region of the array.

[0083] S113. When the seismic source is on the right side of the array, the initialization regions are all concentrated in the right side of the array. However, as long as the seismic source is on the right side of the array, there is a common good initialization region in its initialization region, that is, the initialization regions are all concentrated in the right side of the array.

[0084] In a specific implementation, as a preferred embodiment of the present invention, step S12 specifically includes: investigating whether the distribution of function values ​​is a complex space, resulting in multiple extreme points that cause the algorithm to get stuck in a local optimum and fail to locate accurately; exploring the relationship between the spatial distribution of the target function values ​​and the spatial distribution of the initial point, and in the case of an unknown seismic source, deriving the region of the initial point based on the information of the target function.

[0085] Based on the similarity between the two-dimensional and three-dimensional objective functions, the spatial distribution pattern of the two-dimensional objective function is analyzed:

[0086] The function values ​​do not exhibit multiple distinct extreme regions in the overall space, and their spatial distribution remains spatially dependent on the array position, showing a clear left-right distribution pattern. Furthermore, the distribution of the objective function is somewhat unique; the changes in the objective function are relatively small in the regions on either side of the array.

[0087] The relationship between the spatial distribution of the objective function values ​​and the initial point distribution was explored as follows:

[0088] S121. When the actual microseismic event is located in the left region of the array, the objective function value shows a characteristic of being smaller on the left and larger on the right. The distribution of the objective function value in the regions on both sides of the array is close to a plane, indicating that the change of the objective function value in the regions on both sides of the array is smaller than that in the middle of the array. The distribution characteristics of the objective function explain why the location of the microseismic event outside the array is inaccurate.

[0089] S122. When the real microseismic event is located in the middle of the array, the target value shows a distribution pattern of large on both sides and small on the right and in the middle. The target function value on both sides changes little, while the target function value in the middle of the array changes significantly.

[0090] S123. When the real microseismic event is located in the right region of the array, the distribution pattern of the objective function is similar to that when it is on the left side of the array, but the distribution pattern in the positive and negative directions of the X-axis is opposite, showing the characteristic of being larger on the left and smaller on the right.

[0091] In a specific implementation, as a preferred embodiment of the present invention, step S13 specifically includes: to further analyze why improper selection of the initial point can lead to incorrect positioning, different initial points are selected for iterative optimization, and the optimization trajectories of different initial points are compared. The comparison yields:

[0092] At different initial points, the corresponding loss functions converge and satisfy the stopping condition. However, the localization results differ depending on the initial point. A good initial point can locate the true coordinates of the microseismic event, while a poor initial point selection will prevent the accurate location. By comparing the locations of these two initial points, it can be seen that the objective function value changes more significantly at the correctly located initial point, meaning the gradient distribution of the objective function is larger. Conversely, the gradient distribution of the objective function is smaller at the initial point that fails to locate the microseismic event. Therefore, it can be concluded that there is a clear correlation between whether the initial point can locate the true location and the gradient distribution of the objective function. This also explains why, even though the function value does not have multiple obvious extreme regions in the overall space, the location still cannot be accurately determined.

[0093] In a specific implementation, as a preferred embodiment of the present invention, step S14 specifically includes: obtaining the relationship between the initialization region and the gradient based on the contour map of the earthquake source location, the gradient corresponding to the objective function, and the function value;

[0094] S141. If the negative gradient direction points to the left, then the initialization region is located in the left region of the sensor array.

[0095] S142. If the negative gradient direction points to the middle, then the initialization region is located in the middle of the sensor array.

[0096] S143. If the negative gradient direction points to the right, then the initialization region is located in the right region of the sensor array.

[0097] In a specific implementation, as a preferred embodiment of the present invention, step S15 specifically includes: calculating the negative gradient direction at the midpoint of each row of sensors, and taking the side that all points to as the initialization region; the rules for determining the initialization region are as follows:

[0098] Let the angles between the negative gradient direction at the midpoint of the two rows of sensors and the x-axis be α1 and α2, respectively. Then we have:

[0099] S151, Satisfaction Formula The initialization region is located in the left region of the sensor array;

[0100] S152, Satisfaction Formula The initialization area is located in the middle of the sensor array;

[0101] S153, Satisfaction Formula The initialization region is located in the right region of the sensor array.

[0102] In a specific implementation, as a preferred embodiment of the present invention, step S16 specifically includes: initializing a target point within the initialization region; based on the principle that the gradient value is smaller the closer to the minimum value, the initial point is obtained by weighting the coordinates of the four vertices of the region; the calculation formula for the initial point is:

[0103]

[0104] In the formula, Represents weight, d i represents the gradient magnitude, and n represents the four vertices of the region.

[0105] In a specific implementation, as a preferred embodiment of the present invention, step S2 specifically includes:

[0106] S21. Determine the threshold based on the maximum possible distance;

[0107] S22. Based on the maximum threshold, remove data;

[0108] S23. Adaptive weighting based on the two-wave theorem;

[0109] S24. Using an adaptive weighted objective function, the location of the microseismic event is obtained through an optimization solution method.

[0110] In a specific implementation, as a preferred embodiment of the present invention, step S21 specifically includes: based on the characteristics of sensor arrangement in deep-buried tunnel engineering, proposing different data filtering thresholds for sensors in the same row and different rows:

[0111] S211. Assume the distance between two rows of sensors is a, and the distances between adjacent sensors in the same row are b and c.

[0112] S212. Different rows of sensors: For any seismic source, the time difference Δt between the P-waves received by any two sensors in different rows of sensors. ij All less than T different_max ,Right now In the formula, S i S′ j This represents the distance between any two sensors in different rows;

[0113] S213. Same row of sensors: For any seismic source, the time difference Δt between the P-waves received by any two sensors in the same row of sensors.ij All less than T same_max ,Right now In the formula, S i S j This represents the distance between any two sensors in the same row.

[0114] This invention will T different_max and T same_max As a threshold standard for filtering out data from different rows of sensors and data from the same row of sensors, and taking into account the actual data situation of the field when applied to the field, two different threshold coefficients α and β are given. This allows for better elimination of significant and highly interfering values, and retains as much data as possible to avoid loss of positioning accuracy due to missing data.

[0115] In a specific implementation, as a preferred embodiment of the present invention, step S22 specifically includes: after deriving the maximum threshold, using a voting algorithm to remove data, such as... Figure 2 As shown, the specific operation process is as follows:

[0116] S221. Divide the sensor data into the same row and different rows, select any two sensors in turn and determine their positions to determine the maximum threshold, and calculate the arrival time difference between the two sensors.

[0117] S222. If the time difference is less than the maximum threshold, the credit value is incremented by 1. Repeat this step until all sensor combinations are traversed.

[0118] S223. Rank all sensors according to their credit scores and remove the 20% of sensors with the lowest credit scores. The remaining sensors will be retained for subsequent processing.

[0119] In a specific implementation, as a preferred embodiment of the present invention, step S23 specifically includes: obtaining the distance based on the two-wavelength theorem using the time difference between P and S, and using the distance to assist in data quality evaluation. The specific operation is as follows:

[0120] S231. Improve the objective function to Using this weighted adaptive objective function, the microseismic event location results are obtained through an optimized solution method; where... This serves as the basis for adaptive weighting of the data; the formula for calculating Δd is: Δd=|d1-d2|, where d1 and d2 represent the distances between the two sensors and the microseismic event, ΔD xyz The third edge representing the spatial relationship between the sensor and microseismic events, such as Figure 3 As shown.

[0121] S232, According to the spatial relation theorem, we obtain Therefore, the data filtering rules are set as follows: The data is unreliable, and The larger the ratio, the less reliable the data.

[0122] 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 them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive localization method for microseismic events in deeply buried tunnels, characterized in that, include: The optimization initialization mechanism based on the target gradient selects a good initial point, including: Analyze the distribution pattern of the initial points; Analyze the spatial distribution of the objective function; Compare the optimization trajectories with different initial points; The study investigates the relationship between the initialization region and the gradient, including: based on the distribution of seismic sources, the gradient corresponding to the objective function, and the contour map of the function value, summarizing the patterns to obtain the relationship between the initialization region and the gradient. If the negative gradient direction points to the left, the initialization region is located in the left region of the sensor array; If the negative gradient direction points to the middle, then the initialization region is located in the middle of the sensor array; If the negative gradient direction points to the right, then the initialization region is located in the right region of the sensor array; Determine the initialization region, including: calculating the negative gradient direction at the midpoint of each row of sensors, and taking the side that all points to as the initialization region; Determine the initial point; By using a data assessment strategy, low-quality data affected by the engineering environment is processed to identify dangerous areas for excavation activities and accurately predict rockburst locations.

2. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 1, characterized in that, The analysis of the distribution patterns of the initial points includes: classifying the earthquake source distribution areas according to the relative positions of the sensor array and the earthquake source, dividing them into three categories: earthquake source on the left side of the array, earthquake source in the middle of the array, and earthquake source on the right side of the array; and analyzing the distribution patterns of the initial points under different earthquake source distribution areas. When the seismic source is on the left side of the array, the initialization regions are all concentrated in the left side of the array. When the seismic source is in the middle of the array, the initialization area is concentrated in the middle of the array; When the seismic source is on the right side of the array, the initialization region is concentrated in the right side region of the array.

3. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 1, characterized in that, The analysis of the spatial distribution of the objective function includes: deriving the two-dimensional pattern from the similarity between the spatial value distribution of the three-dimensional objective function on the z-axis and the spatial distribution of the two-dimensional objective function; and analyzing the relationship between the spatial distribution of the two-dimensional objective function values ​​and the initial point distribution. When the actual microseismic event is located in the left region of the array, the objective function value shows a characteristic of being smaller on the left and larger on the right, and the objective function value on both sides changes little, while the objective function value in the middle of the array changes significantly. When a real microseismic event is located in the middle of the array, the objective function value exhibits a characteristic of being large on both sides and small on the right side and small in the middle. Furthermore, the objective function value on both sides changes little, while the objective function value in the middle of the array changes significantly. When a real microseismic event is located in the right region of the array, the objective function value exhibits a characteristic of being larger on the left and smaller on the right, with the objective function values ​​on both sides changing little, while the objective function value in the middle of the array changes significantly.

4. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 1, characterized in that, The comparison of the optimization trajectories at different initial points includes: Different initial points were selected for iterative optimization. By comparing the optimization trajectories of different initial points, it was found that the positioning results of different initial points were different. Furthermore, the gradient of the objective function of the initial point that could be correctly positioned was large, while the gradient of the objective function of the initial point that could not be correctly positioned was small. Therefore, further optimization of the selection of initial points can be achieved by looking for patterns in the gradients.

5. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 1, characterized in that, The rules for determining the initialization region are as follows: Let the negative gradient direction at the midpoint of the two rows of sensors be parallel to... The angles of the axes are respectively , Then we have: Satisfaction The initialization region is located in the left region of the sensor array; Satisfaction The initialization area is located in the middle of the sensor array; Satisfaction The initialization region is located in the right region of the sensor array.

6. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 1, characterized in that, Determining the initial point includes: initializing the target point within the initialization region; based on the principle that the gradient value decreases as it approaches the minimum, the initial point is obtained by weighting the coordinates of the four vertices of the region; the formula for calculating the initial point is: In the formula, Represents weight, Represents the gradient magnitude. n The four vertices representing the region.

7. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 1, characterized in that, The aforementioned data evaluation strategy involves eliminating low-quality data affected by the engineering environment and formulating a weighting strategy, including: Determine the threshold based on the maximum possible distance; Data is removed based on the maximum threshold. Based on the two-wave theorem, adaptive weighting; The location of microseismic events is obtained by using an adaptive weighted objective function and an optimization solution method.

8. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 7, characterized in that, The threshold determination based on the maximum possible distance includes: proposing different data rejection thresholds for sensors in the same row and different rows, based on the characteristics of sensor arrangement in deep-buried tunnel engineering. Assume the distance between two rows of sensors is *a*, and the distance between adjacent sensors in the same row is *i*. b and c , Different rows of sensors: For any seismic source, the time difference between the P-wave arrival received by any two sensors in different rows of sensors. All less than ,Right now In the formula, This represents the distance between any two sensors in different rows; Same row of sensors: For any seismic source, the signals received by any two sensors in the same row are... Wave to time difference All less than ,Right now In the formula, This represents the distance between any two sensors in the same row.

9. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 7, characterized in that, The data removal based on the maximum threshold includes: after deriving the maximum threshold, using a voting algorithm to remove data. The specific operation process is as follows: The sensor data is divided into the same row and different rows. Any two sensors are selected in turn and their positions are determined to determine the maximum threshold. The arrival time difference between the two sensors is calculated. If the time difference is less than the maximum threshold, the credit value is incremented by 1, and this step is repeated until all sensor combinations are traversed. All sensors are ranked according to their credit scores, and the bottom 20% of sensors are removed. The remaining sensors are retained for further processing.

10. The adaptive localization method for microseismic events in deep-buried tunnels according to claim 7, characterized in that, The adaptive weighting based on the two-wavelength theorem includes: obtaining the distance using the time difference between P and S phases based on the two-wavelength theorem, and using the distance to assist in data quality evaluation. The specific operation is as follows: The objective function is improved into a weight-adaptive objective function. Using a weighted adaptive objective function, the microseismic event location results are obtained through an optimization solution method; among them, As the basis for adaptive weighting of data, the spatial relationship theorem is used to obtain... Therefore, the data filtering rules are set as follows: The data is unreliable, and The larger the ratio, the less reliable the data; in the formula The calculation formula is: , and This represents the distance between the two sensors and the microseismic event. This represents the third edge in the spatial relationship between the sensor and the microseismic event.

Images