Dynamic identification analysis method and system for earth-rock blasting

Abstract

The invention discloses a dynamic recognition and analysis method and system for earth-rock blasting, and belongs to the technical field of earth-rock blasting. A plurality of sensors are arranged in a blasting area, sound wave signals and displacement data generated in the blasting process are collected, and time sequence synchronization processing is carried out to construct a unified data time axis; preprocessing the multi-source heterogeneous data, and extracting a high-frequency resonance duration characteristic and a micro-amplitude reverse displacement trend characteristic; constructing a blasting response state model based on the extracted features, identifying the response state of the blasting effect and performing grading evaluation; the recognition result is compared with an existing geological model and design parameters, whether blasting reaches an expected effect or not is judged, and early warning information is output; the judgment result is visually displayed and subjected to data storage for risk early warning and blasting parameter optimization in the construction process, intelligent identification and dynamic analysis of the blasting effect are achieved, and the safety and scientificity of earth-rock blasting construction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of earthwork blasting, and in particular to a dynamic identification and analysis method and system for earthwork blasting. Background Art

[0002] Dynamic identification and analysis of earthwork blasting involves acquiring real-time, multi-source data such as vibration, sound waves, displacement, and crack expansion generated by blasting during blasting operations through sensors, image processing, and signal analysis. This data is then dynamically monitored, identified, and analyzed using algorithms to assess blasting effectiveness, determine rock mass response characteristics, and identify abnormal conditions, thereby enabling intelligent perception and refined control of the blasting process. This approach transcends the limitations of traditional static empirical judgment and helps improve the safety and efficiency of blasting operations.

[0003] The existing technology has the following shortcomings:

[0004] During highway construction on steep slopes in mountainous areas, the deployment of monitoring equipment is difficult due to terrain constraints, making it difficult to capture the rock mass response characteristics at the moment of blasting using traditional methods. During the construction of plateau tunnel exit sections, if rock mass loosening caused by blasting is not promptly identified, it can cause delayed slippage of exposed rock blocks within minutes of blasting, crushing the newly supported slope structure and causing significant economic losses. Summary of the Invention

[0005] The purpose of the present invention is to provide a dynamic identification and analysis method and system for earthwork blasting to solve the shortcomings of the background technology.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a dynamic identification and analysis method for earthwork blasting, comprising:

[0007] Various sensors are deployed in the blasting area to collect acoustic signals and displacement data generated during the blasting process. The collected multi-source heterogeneous data is then synchronized and processed to build a unified data timeline.

[0008] After preprocessing the multi-source heterogeneous data, the high-frequency resonance duration characteristics in the blasting acoustic wave signal and the slight reverse displacement trend characteristics in the displacement data are extracted;

[0009] Based on the high-frequency resonance duration characteristics and the slight reverse displacement trend characteristics, a blasting response state model is constructed for the response state identification and graded evaluation of blasting effects.

[0010] Compare the identification results with the existing geological model and design parameters to determine whether the blasting has achieved the expected effect and output early warning information;

[0011] The judgment results are visualized and stored in the blasting database for risk warning and blasting parameter optimization during the construction process.

[0012] Preferably, the deployed sensors include acoustic emission sensors and displacement monitoring devices, the acoustic emission sensors are used to collect high-frequency sound wave signals generated during the blasting process, and the displacement monitoring devices include laser rangefinders, displacement meters or total stations, which are used to collect micro-displacement changes of the rock mass.

[0013] Preferably, the extraction of high-frequency resonance duration characteristics from the blasting sound wave signal includes: performing Fourier transform on the sound wave signal to obtain its spectral distribution; identifying energy peaks within a preset high-frequency threshold range, and counting the duration during which the sound energy density in the frequency band exceeds the set threshold as the high-frequency resonance duration.

[0014] Preferably, the high-frequency resonance duration feature in the extracted sound wave signal is analyzed to generate a high-frequency resonance duration abnormal value, and the generation method is:

[0015] The duration of high-frequency resonance of each burst is recorded as time series data: X = {x1, x2, x3, ..., x n}; where x n Represents the duration of the high-frequency resonance extracted from the nth burst; set a sliding window of length w, and for the tth data point, calculate the mean and standard deviation within the sliding window:

[0016] Where x i is the duration of high-frequency resonance extracted from the i-th blast, μ t is the mean duration of high-frequency resonance, σ t is the standard deviation of the high-frequency resonance duration, for the current data point x t Calculate its Z-Score value: A Z-Score threshold is set, and when it exceeds the threshold, it is considered as an abnormal value of high-frequency resonance duration.

[0017] Preferably, the extraction of slight reverse displacement trend features in the displacement data includes: setting an analysis window, extracting the displacement change curve within 0 to 5 minutes after blasting; identifying the trend change of displacement direction from positive to negative, and extracting the maximum reverse amplitude and reverse rate; when the reverse amplitude is greater than a preset threshold and the duration exceeds a set time, it is determined to be a slight reverse trend feature.

[0018] Preferably, the trend deviation angle of the slight reverse displacement is generated after analyzing the slight reverse displacement trend characteristics in the extracted displacement data, and the generation method is:

[0019] Select a sliding window of length M from the original displacement sequence to generate multiple time periods: M t ={y t-w+1 ,y t-w+2 ,...,y t}; where y t Represents the displacement data at the t-th time point, M is the window length; for the current window M t Perform linear fitting on the data within to obtain the current trend vector. The fitting model is: i =a t ·t i +b t ; Among them: a t is the regression slope in the current window, indicating the current displacement trend, b t is the intercept;

[0020] From M t-1 Repeat the linear fitting steps within the window to obtain the reference trend slope a ref ; Convert the two regression vectors into two-dimensional vectors and then calculate the angle:

[0021] is the reference trend vector; the trend deviation angle θ is calculated using the formula:

[0022] Preferably, the blasting response state model is constructed based on a polynomial regression model, specifically including:

[0023] The abnormal value of high-frequency resonance duration and the deviation angle of slight reverse displacement trend are normalized and combined into a comprehensive feature vector;

[0024] The polynomial regression model is trained with the comprehensive feature vector as input and the abnormal analysis value label of blast response state identification as output target;

[0025] The model training is completed when the training error converges by minimizing the sum of squared prediction errors as the loss function.

[0026] Preferably, the acquired abnormal analysis value for identifying the response state of the blasting effect is compared with a gradient standard response threshold, the gradient standard response threshold including a first standard response threshold and a second standard response threshold, and the first standard response threshold is smaller than the second standard response threshold, and the abnormal analysis value for identifying the response state of the blasting effect is compared with the first standard response threshold and the second standard response threshold respectively;

[0027] If the abnormal analysis value of the response state identification of the blasting effect is greater than the second standard response threshold, it is judged that there is a serious abnormal response in the blasting event and marked as a high risk level. The blasting parameters are immediately reviewed or the geological structure stability review process is initiated;

[0028] If the abnormal analysis value of the response state identification of the blasting effect is greater than or equal to the first standard response threshold and less than or equal to the second standard response threshold, it is judged that the blasting event has a moderate abnormal response, marked as a medium risk level, recorded and arranged for subsequent monitoring or parameter optimization;

[0029] If the abnormal analysis value of the response state identification of the blasting effect is less than the first standard response threshold, the blasting event response is judged to be within the normal range and marked as a low risk level, and no additional processing is required.

[0030] Preferably, the determining whether the blasting achieves the expected effect includes:

[0031] Compare the prediction results with the target values ​​of the response parameters set in the existing geological model;

[0032] If the response value deviates from the design value and exceeds the preset tolerance range, a risk warning signal is output.

[0033] The present invention also provides a dynamic identification and analysis system for earthwork blasting, comprising a blasting monitoring data acquisition module, a feature extraction module, a response state evaluation module, a design comparison and early warning module, and a visualization display module;

[0034] Blasting monitoring data acquisition module: Various sensors are deployed in the blasting area to collect acoustic signals and displacement data generated during the blasting process. The collected multi-source heterogeneous data is then synchronized in time series to build a unified data timeline.

[0035] Feature extraction module: After preprocessing multi-source heterogeneous data, it extracts the high-frequency resonance duration characteristics in the blasting sound wave signal and the slight reverse displacement trend characteristics in the displacement data;

[0036] Response state assessment module: Based on the high-frequency resonance duration characteristics and the slight reverse displacement trend characteristics, a blasting response state model is constructed for the response state identification and graded assessment of the blasting effect;

[0037] Design comparison and early warning module: compares the identification results with the existing geological model and design parameters to determine whether the blasting has achieved the expected effect and outputs early warning information;

[0038] Visual display module: The judgment results are displayed visually and stored in the blasting database for risk warning and blasting parameter optimization during the construction process.

[0039] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0040] 1. This invention deploys acoustic emission sensors and displacement monitoring devices in the blasting area to collect acoustic signals and rock mass micro-displacement data during the blasting process. It then performs time-series synchronous processing and unified analysis of different types of data, addressing the technical bottlenecks of traditional blasting monitoring, such as data dispersion, difficulty in information fusion, and a lack of dynamic response feature recognition. By extracting key response features, such as high-frequency resonance duration and micro-reverse displacement trends, and combining them with a polynomial regression model to establish a blasting response state recognition system, this system enables quantitative assessment and grading of blasting effect anomalies, effectively improving the accuracy and real-time nature of blasting behavior recognition.

[0041] 2. By comparing the identification results with the existing geological model and blasting design parameters, the present invention can dynamically determine whether the blasting has achieved the expected effect and automatically generate early warning information when an abnormal state is identified. At the same time, the system will visualize the judgment results, graphically reflect the various response parameters and risk levels, and store them in the blasting database, providing reliable data support for risk warnings and blasting parameter optimization during subsequent construction. Overall, the present invention has the significant advantages of intelligent response identification, systematic information processing, and forward-moving risk management and control, and is suitable for earthwork blasting projects under various complex geological conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0043] Figure 1 This is a mind map of the method of the present invention.

[0044] Figure 2 This is a mind map of the system modules of the present invention. DETAILED DESCRIPTION

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0046] Example 1, please refer to Figure 1 As shown, the dynamic identification and analysis method of earthwork blasting described in this embodiment includes:

[0047] Various sensors are deployed in the blasting area to collect acoustic signals and displacement data generated during the blasting process. The collected multi-source heterogeneous data is then synchronized and processed to build a unified data timeline.

[0048] After preprocessing the multi-source heterogeneous data, the high-frequency resonance duration characteristics in the blasting acoustic wave signal and the slight reverse displacement trend characteristics in the displacement data are extracted;

[0049] Based on the high-frequency resonance duration characteristics and the slight reverse displacement trend characteristics, a blasting response state model is constructed for the response state identification and graded evaluation of blasting effects.

[0050] Compare the identification results with the existing geological model and design parameters to determine whether the blasting has achieved the expected effect and output early warning information;

[0051] The judgment results are visualized and stored in the blasting database for risk warning and blasting parameter optimization during the construction process.

[0052] Before earthwork blasting operations begin, key monitoring locations are prioritized based on the terrain characteristics, rock mass structure, and construction requirements of the blasting area. Various types of sensors are deployed in the blasting area, primarily in the following two categories:

[0053] Acoustic signal acquisition equipment: Use highly sensitive acoustic emission sensors or environmental acoustic wave sensors, which are deployed on the interface between the rock surface and the structure around the blasting area. The deployment points should cover potential crack development areas and rebound-sensitive areas. The sensors need to have a high sampling rate (usually above 100kHz) to capture the acoustic wave pulses at the moment of blasting and support real-time data transmission and remote control. The simultaneous deployment of multiple sensor nodes can realize sound source positioning and local resonance identification.

[0054] Displacement monitoring equipment: Use high-resolution laser rangefinders, total stations or high-frequency displacement meters to monitor the tiny displacement changes of the rock mass before, during and after blasting. For areas prone to slip or structural instability, point-to-point monitoring can be performed in combination with attached strain gauges or displacement sensors. The acquisition frequency setting should cover the main displacement response process during the blasting period, and the commonly used sampling interval is between 0.01 and 0.1 seconds.

[0055] Since acoustic signals and displacement data come from different types of sensors, with different sampling frequencies, heterogeneous data formats, and different triggering mechanisms, the raw data need to be synchronized and processed uniformly. The specific steps are as follows:

[0056] All sensor devices are connected to the same central data acquisition system or synchronized through the GPS timing system to ensure that each data record is accompanied by a unified timestamp;

[0057] During the acquisition process, the acoustic wave data and displacement data are cached separately, and preliminary preprocessing is completed at the same time, such as denoising, signal enhancement, and zero drift correction;

[0058] Convert the acoustic wave data (continuous time domain waveform) and displacement data (interval scalar data) into a unified data structure, usually stored in a time series format. The format includes fields such as timestamp, sensor ID, location number, signal value, and unit.

[0059] Establish a standardized data timeline with a uniform time interval (e.g., every 10 milliseconds); align the sensor data to this timeline, and fill in missing data through interpolation or linear extrapolation. This step ensures that data in all dimensions are analyzed on the same time basis.

[0060] The synchronized multi-source data will serve as the input data source for subsequent feature extraction and pattern recognition. It has temporal consistency and can realize dynamic process reconstruction and blasting behavior identification.

[0061] After completing the unified time axis construction and preprocessing of the blasting acoustic wave signal and displacement data, the key features are extracted using the following method:

[0062] Extracting high-frequency resonance duration characteristics from acoustic signals: Perform a Fourier transform (FFT) on the preprocessed blasting acoustic signal to obtain its spectral distribution. Resonance effects during blasting are typically most pronounced in the high-frequency range between 3 and 15 kHz. A high-frequency threshold range, such as 8 to 12 kHz, is set. The peak sound pressure energy within this frequency range is identified and its duration is observed.

[0063] The duration of time during which the acoustic energy density in this high-frequency band remains above a set energy threshold (e.g., 20% of the total acoustic energy density) is recorded and defined as the "high-frequency resonance duration" parameter. This parameter reflects the resonant response behavior of local voids or loose structures in the rock mass. This duration is used as a structural response feature of the acoustic signal and is associated with the corresponding blast location and time stamp to facilitate subsequent model fusion analysis.

[0064] Extract the characteristics of slight reverse displacement trends in displacement data: Extract the original displacement change curve in a short time window after the blast (such as within 0-5 minutes), remove the drastic change section, and retain the slight response interval. Calculate the displacement increments for consecutive time periods, focusing on the area where the displacement suddenly changes from positive (away from the explosion center) to negative (toward the explosion center or a slight rebound in the opposite direction). Set a minimum reverse displacement threshold (for example, ≥0.3mm) and a duration requirement (for example, ≥20 seconds). When the conditions are met, it is identified as a "slight reverse displacement trend" event, which is commonly seen in rock rebound, stress adjustment, or potential sliding precursors after blasting. Extract parameters such as the reverse displacement change rate (mm / min), maximum reverse amplitude, and occurrence time as feature outputs to support post-blasting stability analysis.

[0065] The duration of high-frequency resonance can sensitively reflect the local resonance effect during the penetration of the blasting wave, and has unique advantages in evaluating the internal structural integrity of the rock mass; the trend of slight reverse displacement is a feature that is easily overlooked in conventional monitoring but has great early warning value. It often appears before slope instability or surrounding rock rebound, and is an important basis for judging dynamic stability.

[0066] After analyzing the high-frequency resonance duration characteristics in the extracted sound wave signal, a high-frequency resonance duration anomaly value is generated. The generation method is:

[0067] The duration of high-frequency resonance of each burst is recorded as time series data: X = {x1, x2, x3, ..., x n}; where x n Indicates the duration of high-frequency resonance extracted from the nth burst (unit: ms or s). Set a sliding window of length w (e.g., w = 10), that is, take 10 consecutive data as a subsample interval each time to perform local statistical analysis.

[0068] For the t-th data point (when t>w), calculate the mean and standard deviation within the sliding window:

[0069] Where xi is the duration of high-frequency resonance extracted from the ith blast, μ t is the mean duration of high-frequency resonance, σ t is the standard deviation of the high-frequency resonance duration, for the current data point x t Calculate its Z-Score value: Set the Z-Score threshold (e.g. |Z t |>2.5), when it exceeds the threshold, the point is considered to be an abnormal value of high-frequency resonance duration.

[0070] It should be noted that in actual deployment, it is recommended that the Z threshold be set dynamically (for example, in combination with the 3σ rule or IQR).

[0071] After analyzing the trend characteristics of the slightly reverse displacement in the extracted displacement data, the trend deviation angle of the slightly reverse displacement is generated. The generation method is:

[0072] Select a sliding window of length M from the original displacement sequence to generate multiple time periods: M t ={y t-w+1 ,y t-w+2 ,...,y t}; where y t represents the displacement data at the t-th time point, and M is the window length (e.g., 10 data points).

[0073] For the current window M t Perform linear fitting on the data to obtain the current trend vector (regression slope): The fitting model is: y i =a t ·t i +b t ; Among them: a t is the regression slope in the current window, indicating the current displacement trend, which can be calculated using the least squares method; b t is the intercept and is not used for deviation calculation.

[0074] From an earlier window (such as M t-1 ) Repeat the linear fitting steps to obtain the reference trend slope a ref ; Convert the two regression vectors into two-dimensional vectors (unit vectors) and then calculate the angle:

[0075] is the reference trend vector; the trend deviation angle θ is calculated using the formula: If θ is less than or equal to 15 degrees: it is considered as trend continuation with no obvious deviation; if θ is between 15 and 60 degrees: there may be a change in direction, further observation is required; if θ is greater than or equal to 60 degrees: there is a significant directional deviation, especially when the slope changes from positive to negative (a rebound trend occurs after an explosion), it should be marked as a reverse trend deviation anomaly.

[0076] Based on the high-frequency resonance duration characteristics and the slight reverse displacement trend characteristics, a blasting response state model is constructed for the response state identification and graded evaluation of blasting effects.

[0077] The high-frequency resonance duration anomaly values ​​and trend deviation angles are converted into comprehensive feature vectors, which are used as inputs of a machine learning model. The machine learning model uses each set of comprehensive feature vectors to predict the response state identification anomaly analysis value label of the blasting effect as a prediction target, and uses minimizing the sum of the prediction errors of the response state identification anomaly analysis value labels of all blasting effects as a training target. The machine learning model is trained until the sum of the prediction errors converges, and the model training is stopped. The response state identification anomaly analysis value of the blasting effect is determined based on the model output results. The machine learning model is a polynomial regression model.

[0078] Comparing the acquired abnormal analysis value of the response state identification of the blasting effect with a gradient standard response threshold, wherein the gradient standard response threshold includes a first standard response threshold and a second standard response threshold, and the first standard response threshold is less than the second standard response threshold, and comparing the abnormal analysis value of the response state identification of the blasting effect with the first standard response threshold and the second standard response threshold respectively;

[0079] If the abnormal analysis value of the response state identification of the blasting effect is greater than the second standard response threshold, it is judged that there is a serious abnormal response in the blasting event and marked as a high risk level. It is recommended to immediately review the blasting parameters or start the geological structure stability review process;

[0080] If the abnormal analysis value of the response state identification of the blasting effect is greater than or equal to the first standard response threshold and less than or equal to the second standard response threshold, it is judged that the blasting event has a moderate abnormal response and is marked as a medium risk level. It is recommended to record and arrange subsequent monitoring or parameter optimization;

[0081] If the abnormal analysis value of the response state identification of the blasting effect is less than the first standard response threshold, it is judged that the blasting event response is within the normal range and marked as a low risk level, and no special treatment is required.

[0082] After obtaining the blasting effect response state identification anomaly analysis value Rpred based on comprehensive eigenvector analysis, the identification result is compared with the existing geological model parameters and the design blasting parameter target values ​​in the project to determine whether the actual blasting has achieved the expected effect. The specific steps include:

[0083] The three-dimensional geological model information corresponding to the current blasting area is extracted from the project database, including the stratigraphic structure, joint and fissure distribution, rock mass stability assessment results, etc., as a reference for the basic conditions for the expected blasting response.

[0084] Call the target parameters set in the design phase, such as: blasting vibration control threshold, expected crack control range, allowable displacement limit, designed energy release cycle, etc.

[0085] If the identified outlier value Rpred is greater than the expected response target value, or shows an obvious trend deviation (such as the reverse trend angle is much higher than the reference value), it is determined that the blasting effect has not met expectations; if the predicted value is basically consistent with the design expectation, or falls within the tolerance interval, it is considered to have achieved the design target.

[0086] If the expected results are not achieved or a safety risk trend is identified, the system will generate an early warning message, including: blasting event number and time; key parameters exceeding design targets; possible risk locations (combined with displacement sensor placement); and recommended measures (such as reviewing blasting parameters, adjusting the next charge or drill hole spacing). Real-time warnings can be delivered via the platform interface, mobile terminals, or integrated monitoring systems, supporting graphical visual analysis and decision-making assistance.

[0087] After identifying and judging the blasting effect response status, the system further visualizes the judgment results to improve project operability and decision-making efficiency. It also stores the core data in the blasting database for use in risk warning and blasting parameter optimization during construction. The specific steps are as follows:

[0088] Based on the engineering monitoring platform, a graphical visualization interface is constructed to dynamically display the identification results of each blasting event through multi-dimensional data mapping, including: blasting area location map; high-frequency resonance duration curve; displacement trend angle change map; comprehensive abnormal analysis value bar chart or thermal map; risk level classification color identification (such as green / yellow / red marking).

[0089] Users can click on a blasting point to display the detailed identification results of the event, comparative analysis of design targets, trends of key characteristic parameters, whether an early warning is triggered, and recommended adjustment measures.

[0090] The identification results and analysis labels of each blasting are stored in a structured format in the blasting database. The storage fields include: blasting number and timestamp; blasting location and corresponding sensor number; high-frequency resonance duration abnormality value; slight reverse displacement trend deviation angle; comprehensive abnormality analysis value; judgment level (normal, medium, severe); whether to trigger an early warning (yes / no); and recommended optimization suggestions.

[0091] The blasting database is linked to the risk warning system at the construction site. When new blasting data is written into the database and triggers a high-risk indicator, the system automatically calls the record and generates a construction prompt or alarm message.

[0092] The historical records in the database will serve as the input of the blasting parameter optimization system. Combined with machine learning or empirical regression models, iterative optimization suggestions will be made for parameters such as charge amount, hole spacing, and blasting sequence to achieve intelligent closed-loop adjustment of blasting design.

[0093] Example 2, please refer to Figure 2 As shown, the dynamic identification and analysis system for earthwork blasting described in this embodiment includes a blasting monitoring data acquisition module, a feature extraction module, a response state evaluation module, a design comparison and early warning module, and a visualization display module;

[0094] Blasting monitoring data acquisition module: Various sensors are deployed in the blasting area to collect acoustic signals and displacement data generated during the blasting process. The collected multi-source heterogeneous data is then synchronized in time series to build a unified data timeline.

[0095] Feature extraction module: After preprocessing multi-source heterogeneous data, it extracts the high-frequency resonance duration characteristics in the blasting sound wave signal and the slight reverse displacement trend characteristics in the displacement data;

[0096] Response state assessment module: Based on the high-frequency resonance duration characteristics and the slight reverse displacement trend characteristics, a blasting response state model is constructed for the response state identification and graded assessment of the blasting effect;

[0097] Design comparison and early warning module: compares the identification results with the existing geological model and design parameters to determine whether the blasting has achieved the expected effect and outputs early warning information;

[0098] Visual display module: The judgment results are displayed visually and stored in the blasting database for risk warning and blasting parameter optimization during the construction process.

[0099] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0100] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0101] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0102] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A dynamic identification and analysis method for earthwork blasting, characterized by: include: Various sensors are deployed in the blasting area to collect acoustic signals and displacement data generated during the blasting process. The collected multi-source heterogeneous data is then synchronized and processed to build a unified data timeline. After preprocessing the multi-source heterogeneous data, the high-frequency resonance duration characteristics in the blasting acoustic wave signal and the slight reverse displacement trend characteristics in the displacement data are extracted; Based on the high-frequency resonance duration characteristics and the slight reverse displacement trend characteristics, a blasting response state model is constructed for the response state identification and graded evaluation of blasting effects. Compare the identification results with the existing geological model and design parameters to determine whether the blasting has achieved the expected effect and output early warning information; The judgment results are visualized and stored in the blasting database for risk warning and blasting parameter optimization during the construction process.

2. The dynamic identification and analysis method for earthwork blasting according to claim 1, characterized in that: The deployed sensors include acoustic emission sensors and displacement monitoring devices. The acoustic emission sensors are used to collect high-frequency sound wave signals generated during the blasting process. The displacement monitoring devices include laser rangefinders, displacement meters or total stations for collecting micro-displacement changes of rock masses.

3. The dynamic identification and analysis method for earthwork blasting according to claim 1, characterized in that: The method of extracting the high-frequency resonance duration feature from the blasting sound wave signal includes: performing Fourier transform on the sound wave signal to obtain its spectrum distribution; identifying the energy peak within a preset high-frequency threshold range, and counting the duration during which the sound energy density in the frequency band exceeds the set threshold as the high-frequency resonance duration.

4. The dynamic identification and analysis method for earthwork blasting according to claim 3, characterized in that: After analyzing the high-frequency resonance duration characteristics in the extracted sound wave signal, a high-frequency resonance duration anomaly value is generated. The generation method is: The duration of high-frequency resonance of each burst is recorded as time series data: X = {x1, x2, x3, ..., x n }; where x n Represents the duration of the high-frequency resonance extracted from the nth burst; set a sliding window of length w, and for the tth data point, calculate the mean and standard deviation within the sliding window: Where x i is the duration of high-frequency resonance extracted from the i-th blast, μ t is the mean duration of high-frequency resonance, σ t is the standard deviation of the high-frequency resonance duration, for the current data point x t Calculate its Z-Score value: A Z-Score threshold is set, and when it exceeds the threshold, it is considered as an abnormal value of high-frequency resonance duration.

5. The dynamic identification and analysis method for earthwork blasting according to claim 4 is characterized in that: The method of extracting the slight reverse displacement trend feature in the displacement data includes: setting an analysis window and extracting the displacement change curve within 0 to 5 minutes after the blast; identifying the trend change of the displacement direction from positive to negative, and extracting the maximum reverse amplitude and reverse rate; when the reverse amplitude is greater than a preset threshold and the duration exceeds a set time, it is determined to be a slight reverse trend feature.

6. The dynamic identification and analysis method for earthwork blasting according to claim 5, characterized in that: After analyzing the trend characteristics of the slightly reverse displacement in the extracted displacement data, the trend deviation angle of the slightly reverse displacement is generated. The generation method is: Select a sliding window of length M from the original displacement sequence to generate multiple time periods: M t ={y t-w+1 ,y t-w+2 ,...,y t }; where y t Represents the displacement data at the t-th time point, M is the window length; for the current window M t Perform linear fitting on the data within to obtain the current trend vector. The fitting model is: i =a t ·t i +b t ; Among them: a t is the regression slope in the current window, indicating the current displacement trend, b t is the intercept; From M t-1 Repeat the linear fitting steps within the window to obtain the reference trend slope a ref ; Convert the two regression vectors into two-dimensional vectors and then calculate the angle: is the reference trend vector; Calculate the trend deviation angle θ, the formula is:

7. The dynamic identification and analysis method for earthwork blasting according to claim 6, characterized in that: The blasting response state model is constructed based on a polynomial regression model, specifically including: The abnormal value of high-frequency resonance duration and the deviation angle of slight reverse displacement trend are normalized and combined into a comprehensive feature vector; The polynomial regression model is trained with the comprehensive feature vector as input and the abnormal analysis value label of blast response state identification as output target; The model training is completed when the training error converges by minimizing the sum of squared prediction errors as the loss function.

8. The dynamic identification and analysis method for earthwork blasting according to claim 7, characterized in that: Comparing the acquired abnormal analysis value of the response state identification of the blasting effect with a gradient standard response threshold, wherein the gradient standard response threshold includes a first standard response threshold and a second standard response threshold, and the first standard response threshold is less than the second standard response threshold, and comparing the abnormal analysis value of the response state identification of the blasting effect with the first standard response threshold and the second standard response threshold respectively; If the abnormal analysis value of the response state identification of the blasting effect is greater than the second standard response threshold, it is judged that there is a serious abnormal response in the blasting event and marked as a high risk level. The blasting parameters are immediately reviewed or the geological structure stability review process is initiated; If the abnormal analysis value of the response state identification of the blasting effect is greater than or equal to the first standard response threshold and less than or equal to the second standard response threshold, it is judged that the blasting event has a moderate abnormal response, marked as a medium risk level, recorded and arranged for subsequent monitoring or parameter optimization; If the abnormal analysis value of the response state identification of the blasting effect is less than the first standard response threshold, the blasting event response is judged to be within the normal range and marked as a low risk level, and no additional processing is required.

9. The dynamic identification and analysis method for earthwork blasting according to claim 8, characterized in that: Determining whether the blasting has achieved the expected effect includes: Compare the prediction results with the target values ​​of the response parameters set in the existing geological model; If the response value deviates from the design value and exceeds the preset tolerance range, a risk warning signal is output.

10. A dynamic identification and analysis system for earthwork blasting, used to implement the dynamic identification and analysis method for earthwork blasting according to any one of claims 1 to 9, characterized in that: It includes blasting monitoring data acquisition module, feature extraction module, response status assessment module, design comparison and early warning module, and visualization display module; Blasting monitoring data acquisition module: Various sensors are deployed in the blasting area to collect acoustic signals and displacement data generated during the blasting process. The collected multi-source heterogeneous data is then synchronized in time series to build a unified data timeline. Feature extraction module: After preprocessing multi-source heterogeneous data, it extracts the high-frequency resonance duration characteristics in the blasting sound wave signal and the slight reverse displacement trend characteristics in the displacement data; Response state assessment module: Based on the high-frequency resonance duration characteristics and the slight reverse displacement trend characteristics, a blasting response state model is constructed for the response state identification and graded assessment of the blasting effect; Design comparison and early warning module: compares the identification results with the existing geological model and design parameters to determine whether the blasting has achieved the expected effect and outputs early warning information; Visual display module: The judgment results are displayed visually and stored in the blasting database for risk warning and blasting parameter optimization during the construction process.

Images