Hydrological model dynamic monitoring simulation method and system based on multi-dimensional data

Abstract

This application relates to a method and system for dynamic monitoring and simulation of hydrological models based on multidimensional data. The method includes: decomposing microseismic signals from underground mines to obtain multiple wavelet layers; calculating a first noise level for each wavelet layer based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration; determining a second noise level for each wavelet layer based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers; and filtering each wavelet layer based on the fusion result of the first and second noise levels to reconstruct the microseismic signal from all filtered wavelet layers. This application can improve the accuracy of dynamic monitoring and simulation of hydrological models based on multidimensional data.

Description

Technical Field

[0001] This application relates to the field of data monitoring technology, and in particular to a method and system for dynamic monitoring and simulation of hydrological models based on multidimensional data. Background Technology

[0002] In the process of hydrological dynamic monitoring and simulation based on multi-source data, it is usually necessary to integrate multi-dimensional spatiotemporal information such as long-term underground hydrological observation well data, water inflow data, drainage data, precipitation data, and microseismic monitoring basic data in real time through the coal mine hydrological monitoring and early warning platform. Through numerical simulation technology, the aquifer parameters are inverted by data-driven inversion, the coupling mechanism of seepage field and stress field is corrected in real time, the evolution of water-conducting channels is accurately identified, and precise monitoring, early warning and feedback are carried out for possible coal mine water hazards. However, due to the blasting operations commonly found in coal mines, signals are susceptible to external shocks and pulse interference, which can easily lead to misinterpretations as precursors to large-energy water inrushes. Furthermore, high-frequency electromagnetic interference from underground frequency converters, high-power coal mining machine cables, and electric locomotive overhead lines in coal mines can obscure microseismic signals generated during the expansion of mining-induced fractures in the coal seam floor and the formation of water-conducting channels. Existing technologies are prone to over-filtering, resulting in the loss of signal details and significant errors in subsequent platform input signals. This leads to erroneous water hazard alarms from monitoring and early warning platforms, requiring more costly verification and reducing the efficiency of hydrological dynamic monitoring. Summary of the Invention

[0003] To address the aforementioned technical problems, the purpose of this application is to provide a method and system for dynamic monitoring and simulation of hydrological models based on multidimensional data. The specific technical solution adopted is as follows: Firstly, a method for dynamic monitoring and simulation of hydrological models based on multidimensional data is provided, the method comprising: The microseismic signals in the mine were decomposed to obtain multiple wavelet layers; The first noise level of each wavelet layer is calculated based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer; where the first noise level indicates the degree of impulse interference to the wavelet layer, and the square of the detail coefficients of each wavelet layer is proportional to its corresponding energy. The second noise level of each wavelet layer is determined based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers; the second noise level indicates the degree to which the wavelet layer is affected by high-frequency electromagnetic interference. Based on the fusion results of the first and second noise levels, each wavelet layer is filtered to reconstruct the microseismic signal from all the filtered wavelet layers; the reconstructed microseismic signal is used for dynamic monitoring and simulation of mine water hazards.

[0004] Optionally, the microseismic signal from the mine can be decomposed to obtain multiple wavelet layers, including: Variational mode decomposition was performed on the microseismic signals in the mine to obtain multiple intrinsic mode components; Wavelet decomposition is performed on each intrinsic mode component to obtain multiple wavelet layers in each intrinsic mode component.

[0005] Optionally, based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer, the first noise level of each wavelet layer is calculated, including: The energy concentration of each wavelet layer is obtained based on the energy deviation between the peak value of each wavelet layer and its neighboring detail coefficients; the neighboring detail coefficients of each detail coefficient peak indicate the detail coefficients within a preset time window before and after the occurrence of the detail coefficient peak. The degree of drastic change in the detail coefficients of each wavelet layer is obtained by considering the differences between the peak values ​​of adjacent detail coefficients in each wavelet layer. The energy concentration of each wavelet layer is fused with the degree of drastic change in its detail coefficients to obtain the first noise level of each wavelet layer.

[0006] Optionally, the energy concentration of each wavelet layer is obtained based on the energy deviation between the peak value of each wavelet layer and its neighboring detail values, including: Calculate the ratio of the energy corresponding to the peak value of each detail coefficient in each wavelet layer to the mean of the energy corresponding to multiple detail coefficients in its neighborhood, and obtain the energy deviation corresponding to the peak value of each detail coefficient. The energy concentration of each wavelet layer is obtained by taking the average of the energy deviations corresponding to the peak values ​​of all detail coefficients in each wavelet layer.

[0007] Optionally, the degree of drastic change in the detail coefficients of each wavelet layer is obtained based on the difference between the peak values ​​of adjacent detail coefficients in each wavelet layer, including: Arrange all the detail coefficient peaks in each wavelet layer in chronological order to obtain the detail coefficient peak sequence of each wavelet layer; Calculate the absolute difference between each pair of adjacent detail coefficient peaks in the detail coefficient peak sequence of each wavelet layer to obtain multiple detail coefficient peak difference values; The mean of the peak differences of multiple detail coefficients in each wavelet layer is calculated to obtain the degree of drastic change of the detail coefficients in each wavelet layer.

[0008] Optionally, a second noise level for each wavelet layer is determined based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers, including: Based on the amplitude of each detail coefficient peak in each wavelet layer, the detail coefficient peaks in the detail coefficient peak sequence of each wavelet layer are clustered to obtain the peak cluster and low peak cluster corresponding to each wavelet layer; the peak cluster represents the set of detail coefficient peaks with high amplitude, and the low peak cluster represents the set of detail coefficient peaks with low amplitude. Arrange the peak values ​​of detail coefficients in the peak clusters corresponding to each wavelet layer in chronological order to obtain the peak distribution sequence of each wavelet layer. The second noise level of each wavelet layer is calculated based on the distance between the peak distribution sequence of each wavelet layer and the peak distribution sequences of other wavelet layers under the same intrinsic mode component.

[0009] Optionally, based on the distance between the peak distribution sequence of each wavelet layer and the peak distribution sequences of other wavelet layers under the same intrinsic mode component, a second noise level for each wavelet layer is calculated, including: Calculate the Frescher distance between the peak point distribution sequence of each wavelet layer and the peak point distribution sequence of each other wavelet layer under the same intrinsic mode component to obtain multiple Frescher distances corresponding to each wavelet layer; The mean of multiple Fraser distances corresponding to each wavelet layer is calculated to obtain the second noise level of each wavelet layer.

[0010] Optionally, based on the fusion result of the first noise level and the second noise level, each wavelet layer is filtered to reconstruct the microseismic signal based on all the filtered wavelet layers, including: Calculate the high-frequency variability of each wavelet layer based on the variability of its detail coefficients. Calculate the mean of the high-frequency fluctuations of all wavelet layers in each intrinsic mode component to obtain the feature weight of each intrinsic mode component; The first noise level, the second noise level, and the feature weights of the intrinsic mode components corresponding to each wavelet layer are fused to obtain the comprehensive weight of each wavelet layer. Based on the comprehensive weight of each wavelet layer, the denoising threshold corresponding to each wavelet layer is adjusted to obtain the adjusted denoising threshold. Based on the adjusted denoising threshold, each wavelet layer is filtered to reconstruct the microseismic signal from all the filtered wavelet layers.

[0011] Optionally, the denoising threshold corresponding to each wavelet layer is adjusted according to the comprehensive weight of each wavelet layer to obtain the adjusted denoising threshold, including: The adjusted denoising threshold for each wavelet layer is obtained by multiplying the comprehensive weight of each wavelet layer with the preset reference wavelet threshold of its intrinsic mode component.

[0012] Secondly, a dynamic monitoring and simulation system for hydrological models based on multidimensional data is provided, the system comprising: The decomposition module is used to decompose the microseismic signals in the mine to obtain multiple wavelet layers; The calculation module is used to calculate the first noise level of each wavelet layer based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer; wherein, the first noise level indicates the degree of impulse interference to the wavelet layer, and the square of the detail coefficients of each wavelet layer is proportional to its corresponding energy; The determination module is used to determine the second noise level of each wavelet layer based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers; the second noise level indicates the degree to which the wavelet layer is affected by high-frequency electromagnetic interference. The filtering module is used to filter each wavelet layer based on the fusion result of the first noise level and the second noise level, so as to reconstruct the microseismic signal based on all the filtered wavelet layers; the reconstructed microseismic signal is used for dynamic monitoring and simulation of mine water hazards.

[0013] Based on common knowledge in the field, the above-mentioned preferred conditions can be combined arbitrarily to obtain various preferred embodiments of this application.

[0014] This application offers the following advantages: It calculates a first noise level based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer, and a second noise level based on the similarity of the peak distribution of detail coefficients among wavelet layers. The filtering threshold of each wavelet layer is adaptively adjusted based on the fusion result of these two calculations. This allows for differentiated suppression of both pulse interference from underground blasting operations in coal mines and high-frequency electromagnetic interference caused by high-power equipment. Since the square of the detail coefficients in each wavelet layer is proportional to its corresponding energy, the degree of energy concentration accurately reflects the energy characteristics of pulse interference. Furthermore, the similarity analysis of the peak distribution of detail coefficients among different wavelet layers effectively distinguishes between high-frequency electromagnetic interference and normal microseismic signals. This avoids signal detail loss due to over-filtering or erroneous warnings caused by misjudgment of interference, significantly improving the signal-to-noise ratio and detail fidelity of the reconstructed microseismic signal. This, in turn, enhances the accuracy of dynamic monitoring and simulation of hydrological models based on multidimensional data, and improves the reliability of mine water hazard warnings. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application 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 only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a hydrological model dynamic monitoring and simulation method based on multidimensional data in one embodiment; Figure 2 This is a schematic diagram of the structure of a hydrological model dynamic monitoring and simulation system based on multidimensional data in one embodiment; Figure 3 This is a schematic diagram of the structure of an electronic device in one embodiment. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a hydrological model dynamic monitoring and simulation method and system based on multidimensional data proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0019] The specific scheme of the hydrological model dynamic monitoring and simulation method based on multidimensional data provided in this application is described below with reference to the accompanying drawings. For example... Figure 1 As shown, the method includes: S11. The microseismic signal in the mine is decomposed to obtain multiple wavelet layers.

[0020] This application involves installing water level and temperature sensors and pressure sensors in long-term hydrological observation wells underground in coal mines to collect data. Both sensors are set to sample every 30 seconds to avoid data redundancy caused by frequent data collection. Water level and temperature sensors are installed at standardized water troughs at various water inflow points underground to collect water inflow data, with a sampling frequency of once every 30 seconds. Ultrasonic flow meters are installed at the main drainage pipeline and straight channel sections underground to collect drainage volume data, with a sampling frequency of once every 30 seconds. Rain gauges are installed in open, flat, and unobstructed areas around the mine surface to collect precipitation data, with a sampling frequency of once per minute. Fiber optic microseismic sensors are installed inside the surrounding rock (buried in boreholes) underground to collect microseismic signals. Due to the short duration and wide frequency range of microseismic events, an extremely high sampling rate is required; therefore, the sampling frequency is set to 24kHz. All acquired data is transmitted wirelessly to a backend terminal for preprocessing.

[0021] Considering that if a momentary abnormally high value appears in data such as water level or microseismic activity due to sensor vibration, and it is not detected and removed, it will be identified as a drastic hydrological change and immediately issue a high-level flood alarm. Such false alarms will seriously interfere with production. At the same time, considering the risk of missed reports when there are missing data in the detection data, it is necessary to process the missing value and outlier values ​​of the real-time acquired long-term hydrological observation well data, inflow data, drainage data, and precipitation data respectively. Among them, missing value processing can be performed by any one of the linear interpolation method, spline interpolation method, and nearest neighbor interpolation method. Outlier processing can be performed by any one of the 3-Sigma criterion, IQR (interquartile range) method, and MAD (median absolute deviation) method.

[0022] Considering that blasting operations are common in coal mines, which can lead to signal interference from external shocks and pulses, potentially causing misinterpretations as precursors to large-energy water inrushes, and the high-frequency electromagnetic interference generated by underground frequency converters, high-power coal mining machine cables, and electric locomotive overhead lines in coal mines, the optical fiber microseismic sensor's photoelectric demodulation equipment and its subsequent analog-to-digital conversion circuitry can be affected through spatial electromagnetic coupling. This coupling introduces high-frequency noise into the final acquired digital microseismic signal, masking the microseismic signals generated during the expansion of mining-induced fractures in the coal seam floor. Therefore, it is necessary to filter and denoise the real-time acquired microseismic signals.

[0023] In one embodiment, the microseismic signal from underground mine is decomposed to obtain multiple wavelet layers, including: Variational mode decomposition was performed on the microseismic signals in the mine to obtain multiple intrinsic mode components; Wavelet decomposition is performed on each intrinsic mode component to obtain multiple wavelet layers in each intrinsic mode component.

[0024] A time window is set for the real-time acquired microseismic signals, with each time window lasting 1 second. Feature analysis and noise filtering are performed on the corresponding microseismic signals within each time window. Specifically, the microseismic signals within each time window are used as inputs, and the VMD (Variational Mode Decomposition) algorithm is used to decompose the microseismic signals. The number of intrinsic mode components (IMFs) is set to 7 to avoid decomposing too many IMFs, which would render the scale meaningless. The decomposed IMFs are then output. The VMD algorithm is a well-known technique, and the specific operation steps will not be described in detail here.

[0025] Considering the pulse interference noise that may arise from blasting operations within coal mines, blasting is primarily a transient, high-energy release, exhibiting short-duration impact characteristics. Simultaneously, the oscillations generated by blasting decay rapidly and successively from their peak in the form of complex waves, with each peak lasting only a very short time. Conversely, high-frequency electromagnetic interference generated by underground frequency converters, high-power coal mining machine cables, and electric locomotive overhead lines in coal mine scenarios typically manifests as dense, continuous, and irregular oscillations. Given that these two types of interference usually occur in different frequency bands from low to high (pulse interference from blasting is typically concentrated in the low to mid-frequency range, while high-frequency electromagnetic interference from high-power electrical appliances is typically concentrated in the high-frequency range), an analysis is conducted on the different interference scenarios affecting different intrinsic mode components.

[0026] For each intrinsic mode component (EMC), a wavelet thresholding denoising algorithm is used for decomposition and denoising. The db4 wavelet is used for decomposition, dividing each EMC in each time window into five wavelet layers. Each wavelet layer contains a set of detail coefficients, which characterize the amplitude of the high-frequency components of the signal at various time points within that wavelet layer. By analyzing the distribution characteristics of the detail coefficients in each wavelet layer, feature indicators for identifying impulse interference and high-frequency electromagnetic interference can be extracted.

[0027] S12. Calculate the first noise level of each wavelet layer based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer.

[0028] The first noise level indicates the degree of impulse interference to the wavelet layer, and the square of the detail coefficients of each wavelet layer is proportional to its corresponding energy.

[0029] In one embodiment, the first noise level of each wavelet layer is calculated based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer, including: The energy concentration of each wavelet layer is obtained based on the energy deviation between the peak value of each wavelet layer and its neighboring detail coefficients; the neighboring detail coefficients of each detail coefficient peak indicate the detail coefficients within a preset time window before and after the occurrence of the detail coefficient peak. The degree of drastic change in the detail coefficients of each wavelet layer is obtained by considering the differences between the peak values ​​of adjacent detail coefficients in each wavelet layer. The energy concentration of each wavelet layer is fused with the degree of drastic change in its detail coefficients to obtain the first noise level of each wavelet layer.

[0030] Specifically, the energy concentration of each wavelet layer is obtained based on the energy deviation between the peak value of each wavelet layer and its neighboring detail coefficients, including: Calculate the ratio of the energy corresponding to the peak value of each detail coefficient in each wavelet layer to the mean of the energy corresponding to multiple detail coefficients in its neighborhood, and obtain the energy deviation corresponding to the peak value of each detail coefficient. The energy concentration of each wavelet layer is obtained by taking the average of the energy deviations corresponding to the peak values ​​of all detail coefficients in each wavelet layer.

[0031] Considering that blasting is mainly a transient, high-energy release with complex waves, the microseismic signal, after being decomposed to the wavelet layer, exhibits extremely high energy peaks in the wavelet layer where blasting operations cause severe pulse interference. Furthermore, the energy in the wavelet layer is mainly concentrated in the detail coefficients corresponding to the peaks, while the remaining detail coefficients typically have lower energy, presenting isolated spikes with extremely high amplitudes and extremely narrow widths. In contrast, in the case where microseismic fracture details are normally present, since rock microfractures are a continuous acoustic emission phenomenon with a certain duration, their signal is essentially a decaying sine wave or a similar envelope. After being decomposed to the wavelet layer, it exhibits a detail coefficient distribution characteristic with a gradual change in energy amplitude, a more dispersed energy distribution, and an overall lower kurtosis.

[0032] Taking wavelet layer c of the intrinsic mode component b of the signal corresponding to time window a as an example, the peak values ​​of detail coefficients in wavelet layer c are extracted using a peak extraction method. Specifically, a peak-finding algorithm (such as AMPD peak-finding algorithm, derivative method, or curve fitting method) is used to identify local amplitude maxima points from multiple detail coefficients in the wavelet layer, and these local maxima points are taken as the peak values ​​of detail coefficients. The ratio of the energy corresponding to each detail coefficient peak value to the average energy of all detail coefficients within a preset time window before and after its occurrence is calculated to obtain the energy deviation corresponding to each detail coefficient peak value. The average energy deviation corresponding to all detail coefficient peak values ​​in each wavelet layer is calculated to obtain the energy concentration of each wavelet layer. .

[0033] This application converts detail coefficients into energy, which amplifies the difference between the peak value of the detail coefficients and the background, making the pulse interference characteristics generated by blasting operations more significant. This avoids weak pulses being submerged by background noise due to insignificant amplitude differences, thereby improving the calculation accuracy of the first noise level. Furthermore, the degree of energy concentration reflects the true intensity of the pulse interference, which is consistent with the physical nature of focusing on energy release in flood warning, and helps to accurately identify the precursor signals of water inrush.

[0034] In one embodiment, the degree of drastic change in the detail coefficients of each wavelet layer is obtained based on the difference between the peak values ​​of adjacent detail coefficients in each wavelet layer, including: Arrange all the detail coefficient peaks in each wavelet layer in chronological order to obtain the detail coefficient peak sequence of each wavelet layer; Calculate the absolute difference between each pair of adjacent detail coefficient peaks in the detail coefficient peak sequence of each wavelet layer to obtain multiple detail coefficient peak difference values; The mean of the peak differences of multiple detail coefficients in each wavelet layer is calculated to obtain the degree of drastic change of the detail coefficients in each wavelet layer.

[0035] Arrange all detail coefficient peaks in each wavelet layer in chronological order to obtain a detail coefficient peak sequence for each wavelet layer. Calculate the absolute value of the first-order backward difference corresponding to all detail coefficient peaks in the detail coefficient peak sequence of each wavelet layer to obtain multiple detail coefficient peak difference values. That is, calculate the absolute difference between each pair of adjacent detail coefficient peaks in the detail coefficient peak sequence to obtain multiple detail coefficient peak difference values. Understandably, calculating the difference of detail coefficients in this application is also calculating the difference between the amplitudes of detail coefficients. Calculate the mean of multiple detail coefficient peak difference values ​​in each wavelet layer to obtain the initial detail coefficient change drasticness. Then, normalize the initial detail coefficient change drasticness (using Min-Max normalization to map it to the [0,1] interval, setting the maximum and minimum values ​​to the maximum and minimum values ​​of the initial detail coefficient change drasticness of all initial detail coefficient change drasticness in the IMF component to which the wavelet layer belongs) to obtain the detail coefficient change drasticness of the wavelet layer. This is to eliminate the influence of the energy base of different microseismic events on the judgment results.

[0036] The wavelet layer c corresponds to and The product of these values ​​is used as the initial first noise level corresponding to wavelet layer c. This initial first noise level is then normalized to obtain the final first noise level. Specifically, the initial first noise levels of all wavelet layers within the current time window are obtained, their maximum and minimum values ​​are determined, and the Min-Max normalization method is used again to map the initial first noise level of each wavelet layer to the [0,1] interval, resulting in the final first noise level. A larger first noise level indicates a greater difference between the energy corresponding to the peak of the detail coefficients in the wavelet layer and the energy corresponding to adjacent detail coefficients. Simultaneously, the energy corresponding to the peak of the detail coefficients is much higher than the energy corresponding to the other detail coefficients. This usually means that the wavelet layer may be affected by pulse interference from blasting operations, leading to a greater likelihood of extremely narrow isolated spikes and highly concentrated energy at the peak.

[0037] S13. Determine the second noise level of each wavelet layer based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers.

[0038] The second noise level indicates the degree of high-frequency electromagnetic interference affecting the wavelet layer.

[0039] When the intrinsic modal components may contain high-frequency electromagnetic interference noise generated by high-power instruments and normal microseismic crack details, considering that high-frequency electromagnetic interference in coal mine scenarios mainly comes from high-power frequency converters, large motors, or switching power supplies underground, the interference generated by such devices is usually presented in the form of narrowband continuous wave interference. After decomposition to wavelet layers, it is usually concentrated in a few wavelet layers. At the same time, in the wavelet layers severely affected by high-frequency electromagnetic interference, the peak values ​​of detail coefficients will show a very sharp, dense, and stable amplitude distribution, and the distribution of the peak values ​​of detail coefficients is irregular, while other layers usually do not have a similar distribution. In the case of normal conditions that may contain microseismic crack details, the elastic wave generated by the rock microfracture is a physical singular event. As a result, after wavelet decomposition, the singularity point of a signal will simultaneously produce local maxima on all scaling scales (i.e., all wavelet layers), resulting in similar distributions of peak values ​​of detail coefficients in different wavelet layers.

[0040] Therefore, in one embodiment, the second noise level of each wavelet layer is determined based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers, including: Based on the amplitude of each detail coefficient peak in each wavelet layer, the detail coefficient peaks in the detail coefficient peak sequence of each wavelet layer are clustered to obtain the peak cluster and low peak cluster corresponding to each wavelet layer; the peak cluster represents the set of detail coefficient peaks with high amplitude, and the low peak cluster represents the set of detail coefficient peaks with low amplitude. Arrange the peak values ​​of detail coefficients in the peak clusters corresponding to each wavelet layer in chronological order to obtain the peak distribution sequence of each wavelet layer. The second noise level of each wavelet layer is calculated based on the distance between the peak distribution sequence of each wavelet layer and the peak distribution sequences of other wavelet layers under the same intrinsic mode component.

[0041] Specifically, based on the distance between the peak distribution sequence of each wavelet layer and the peak distribution sequences of other wavelet layers under the same intrinsic mode component, the second noise level of each wavelet layer is calculated, including: Calculate the Frescher distance between the peak point distribution sequence of each wavelet layer and the peak point distribution sequence of each other wavelet layer under the same intrinsic mode component to obtain multiple Frescher distances corresponding to each wavelet layer; The mean of multiple Fraser distances corresponding to each wavelet layer is calculated to obtain the second noise level of each wavelet layer.

[0042] Taking wavelet layer c of the intrinsic mode component b corresponding to time window a as an example, the peak value of multiple detail coefficients in the wavelet layer is first extracted using a peak-finding algorithm to obtain the detail coefficient peak sequence. Considering that the detail coefficient peaks generated by micro-crack features or high-frequency electromagnetic interference are usually much higher than the glitches generated by white noise, the k-means clustering algorithm is used to binary classify the peak points in the detail coefficient peak sequence to obtain peak clusters and low-peak clusters. The peak cluster represents the set of detail coefficient peaks with higher amplitude, and the low-peak cluster represents the set of detail coefficient peaks with lower amplitude. It should be noted that before performing clustering, the number of detail coefficient peaks output by the peak-finding algorithm is first obtained. If the number of detail coefficient peaks is less than 10, it indicates that the fluctuation characteristics in the current intrinsic mode component are not significant. In this case, the subsequent clustering and distance calculation steps are not performed, and the feature weight corresponding to the intrinsic mode component is directly set to 0. If the number of detail coefficient peaks is greater than or equal to 10, the subsequent clustering process is performed based on the above detail coefficient peak data.

[0043] Feature analysis is performed on peak clusters. The peak values ​​of detail coefficients within each peak cluster in each wavelet layer are arranged in chronological order to form the peak distribution sequence of the wavelet layer. If the number of elements in the peak distribution sequence of the current wavelet layer is less than or equal to 5, it indicates that the wavelet layer does not have significant singular fluctuations. In this case, the distance calculation step is skipped, and the Fréchet distance between the wavelet layer and other wavelet layers under the same intrinsic mode component is set to a preset minimum distance constant. Otherwise, based on the occurrence of temporal redistribution or by applying a dynamic time warping algorithm, the peak distribution sequences of each wavelet layer are stretched and aligned on the time axis. The Fréchet distance between the peak distribution sequence of the current wavelet layer and the peak distribution sequences of each other wavelet layer under the same intrinsic mode component is calculated, resulting in multiple Fréchet distances for each wavelet layer. The mean of these multiple Fréchet distances for each wavelet layer is then calculated to obtain the second noise level of the wavelet layer.

[0044] The smaller the calculated second noise level value, the more similar the peak distribution sequence of the current wavelet layer is to the peak distribution sequence of other wavelet layers under the same intrinsic mode component. That is, the peak arrangement characteristics in the wavelet layer are similar to the peak arrangement characteristics of other wavelet layers. At this time, the wavelet layer is more likely to be a normal microseismic signal containing microseismic crack details and is less affected by high-frequency electromagnetic interference.

[0045] The first noise metric quantifies the degree of pulse interference generated by blasting operations, while the second noise metric quantifies the degree of high-frequency electromagnetic interference generated by equipment such as frequency converters and high-power coal mining machine cables. Both metric characterize noise features from two dimensions: energy concentration and interlayer peak distribution similarity. Energy concentration can accurately reflect the unique energy peak characteristics of pulse interference generated by blasting operations, thereby accurately identifying transient strong energy impacts caused by blasting and avoiding misjudging them as precursors to water inrush. Analyzing the similarity of detail coefficient peak distributions between different wavelet layers can effectively distinguish between high-frequency electromagnetic interference generated by equipment such as underground frequency converters and high-power coal mining machine cables and normal microseismic crack signals.

[0046] S14. Based on the fusion result of the first noise level and the second noise level, filter each wavelet layer to reconstruct the microseismic signal based on all the filtered wavelet layers.

[0047] The reconstructed microseismic signals are used for dynamic monitoring and simulation of mine water hazards.

[0048] In one embodiment, each wavelet layer is filtered based on the fusion result of the first noise level and the second noise level to reconstruct the microseismic signal based on all filtered wavelet layers, including: Calculate the high-frequency variability of each wavelet layer based on the variability of its detail coefficients. Calculate the mean of the high-frequency fluctuations of all wavelet layers in each intrinsic mode component to obtain the feature weight of each intrinsic mode component; The first noise level, the second noise level, and the feature weights of the intrinsic mode components corresponding to each wavelet layer are fused to obtain the comprehensive weight of each wavelet layer. Based on the comprehensive weight of each wavelet layer, the denoising threshold corresponding to each wavelet layer is adjusted to obtain the adjusted denoising threshold. Based on the adjusted denoising threshold, each wavelet layer is filtered to reconstruct the microseismic signal from all the filtered wavelet layers.

[0049] Specifically, based on the comprehensive weight of each wavelet layer, the denoising threshold corresponding to each wavelet layer is adjusted to obtain the adjusted denoising threshold, including: The adjusted denoising threshold for each wavelet layer is obtained by multiplying the comprehensive weight of each wavelet layer with the preset reference wavelet threshold of its intrinsic mode component.

[0050] Considering the potential pulse interference noise caused by blasting and high-frequency electromagnetic interference noise generated by high-power instruments, a larger denoising threshold should be used for the corresponding wavelet layer to ensure that the signal can be effectively filtered out of interference. For wavelet layers that may contain details of normal microseismic cracks, a smaller denoising threshold should be used to preserve effective signal details.

[0051] Based on this, the feature weight C for each intrinsic mode component is first determined. Specifically, for each wavelet layer of each intrinsic mode component, all detail coefficients within that wavelet layer are extracted, and the variance or standard deviation of these detail coefficients is calculated as a characterization of the high-frequency fluctuation level of that wavelet layer. The mean or weighted average of the variances or standard deviations of all wavelet layers in each intrinsic mode component is calculated to obtain the overall high-frequency fluctuation level corresponding to each intrinsic mode component. This level is then normalized to obtain the feature weight C of the intrinsic mode component. The feature weight C is used to characterize the severity of high-frequency electromagnetic interference affecting the intrinsic mode component as a whole. The larger the C value, the more severe the high-frequency electromagnetic interference affecting the corresponding intrinsic mode component.

[0052] The second noise level of all wavelet layers in each intrinsic mode component is normalized. Specifically, the Min-Max normalization method is used to map the second noise level of all wavelet layers in all intrinsic mode components within the current time window to the interval [0,1], where the maximum and minimum values ​​are set to the maximum and minimum values ​​of all second noise levels within the time window, respectively. The feature weight of the intrinsic mode component to which the wavelet layer belongs is obtained and denoted as C. The formula for calculating the comprehensive weight of each wavelet layer is: Comprehensive weight = max(first noise level, normalized second noise level) + preset base threshold, where the base threshold is set according to the actual situation.

[0053] A fixed threshold rule (such as sqtwolog) can be used to obtain the reference wavelet threshold corresponding to each intrinsic mode component. The comprehensive weight of each wavelet layer is multiplied by the reference wavelet threshold of its corresponding intrinsic mode component to obtain the adjusted denoising threshold for each wavelet layer. For wavelet layers with severe impulse interference or high-frequency electromagnetic interference, the comprehensive weight is larger, and the adjusted denoising threshold is increased accordingly, thereby effectively filtering out noise. For wavelet layers that may contain details of normal microseismic cracks, the comprehensive weight is smaller, and the adjusted denoising threshold is decreased accordingly, thereby preserving signal details.

[0054] Real-time acquired microseismic signals are used as input, and a variational mode decomposition algorithm is applied to decompose the signals into multiple intrinsic mode components (EMS). Utilizing the parallel computing capabilities of the server, wavelet decomposition is performed on each EMS to obtain multiple wavelet layers within each EMS, and these wavelet layers are analyzed synchronously and in parallel. The wavelet threshold is optimized and adjusted by considering the pulse interference characteristics generated by blasting operations and the high-frequency electromagnetic interference characteristics generated by underground frequency converters, high-power coal mining machine cables, and electric locomotive overhead lines. A soft-threshold denoising method is employed to filter and denoise each wavelet layer of each EMS, and the filtered signals are reconstructed to output the filtered microseismic signal.

[0055] Preprocessed long-term hydrological observation well data, inflow data, drainage data, precipitation data, and filtered microseismic signals are used as inputs. The corresponding modules of the coal mine hydrological monitoring and early warning platform are then used to integrate and fuse the acquired multidimensional data. The platform provides a visual output of the corresponding dynamic hydrological monitoring status, and different modules within the platform allow for visualized observation of this monitoring data, enabling dynamic monitoring and simulation of the hydrological model based on multidimensional data.

[0056] This application combines the distinguishing features of pulse interference noise, high-frequency electromagnetic interference noise, and normal microseismic crack details, and performs feature weighting on each intrinsic mode component obtained after variational mode decomposition of the microseismic signal. For each wavelet layer obtained after wavelet decomposition of each intrinsic mode component, the wavelet threshold of each wavelet layer is optimized and adjusted based on the waveform and energy distribution characteristics of pulse interference noise generated by blasting operations, as well as the waveform and energy distribution characteristics of high-frequency electromagnetic interference noise generated by underground frequency converters, high-power coal mining machine cables, and electric locomotive overhead lines. This allows for targeted filtering of noise interference while preserving detailed features such as crack propagation in the microseismic signal.

[0057] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0058] This application also provides a dynamic monitoring and simulation system for hydrological models based on multidimensional data, such as... Figure 2 As shown, the system includes: Decomposition module 21 is used to decompose the microseismic signal in the mine to obtain multiple wavelet layers; The calculation module 22 is used to calculate the first noise level of each wavelet layer based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer; wherein, the first noise level indicates the degree of impulse interference to the wavelet layer, and the square of the detail coefficients of each wavelet layer is proportional to its corresponding energy. The determination module 23 is used to determine the second noise level of each wavelet layer based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers; the second noise level indicates the degree to which the wavelet layer is affected by high-frequency electromagnetic interference. The filtering module 24 is used to filter each wavelet layer based on the fusion result of the first noise level and the second noise level, so as to reconstruct the microseismic signal based on all the filtered wavelet layers; the reconstructed microseismic signal is used for dynamic monitoring and simulation of mine water hazards.

[0059] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs.

[0060] Figure 3 This is a schematic diagram of the structure of an electronic device according to an example embodiment of this application. The electronic device includes a memory, a processor, and a computer program stored in the memory and used to run on the processor. When the processor executes the computer program, it implements the method described in any of the above embodiments. Figure 3 The electronic device 30 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0061] like Figure 3 As shown, the electronic device 30 can be manifested as a general-purpose computing device, such as a server device. The components of the electronic device 30 may include, but are not limited to: at least one processor 31, at least one memory 32, and a bus 33 connecting different system components (including memory 32 and processor 31).

[0062] Bus 33 includes a data bus, an address bus, and a control bus.

[0063] The memory 32 may include volatile memory, such as random access memory 321 and / or cache memory 322, and may further include read-only memory 323.

[0064] The memory 32 may also include a program tool 325 (or utility) having a set (at least one) program module 324, such program module 324 including but not limited to: an operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0065] The processor 31 executes various functional applications and data processing, such as the methods provided in any of the above embodiments, by running computer programs stored in the memory 32.

[0066] Electronic device 30 can also communicate with one or more external devices 34 (e.g., keyboard, pointing device, etc.). This communication can be made through input / output interface 35. Furthermore, electronic device 30 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public network, such as the Internet) via network adapter 36. As shown, network adapter 36 communicates with other modules of electronic device 30 via bus 33. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with electronic device 30, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.

[0067] It should be noted that although several units / modules or sub-units / modules of the electronic device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.

[0068] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided in any of the above embodiments.

[0069] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0070] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0071] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the method described in any of the above embodiments.

[0072] The program code for executing the computer program product of this application can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0073] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0074] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application.

[0075] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A dynamic monitoring and simulation method for hydrological models based on multidimensional data, characterized in that, The method includes: The microseismic signals in the mine were decomposed to obtain multiple wavelet layers; The first noise level of each wavelet layer is calculated based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer; where the first noise level indicates the degree of impulse interference to the wavelet layer, and the square of the detail coefficients of each wavelet layer is proportional to its corresponding energy. The second noise level of each wavelet layer is determined based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers; the second noise level indicates the degree to which the wavelet layer is affected by high-frequency electromagnetic interference. Based on the fusion results of the first and second noise levels, each wavelet layer is filtered to reconstruct the microseismic signal from all the filtered wavelet layers; the reconstructed microseismic signal is used for dynamic monitoring and simulation of mine water hazards.

2. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 1, characterized in that, The microseismic signal from the mine is decomposed to obtain multiple wavelet layers, including: Variational mode decomposition was performed on the microseismic signals in the mine to obtain multiple intrinsic mode components; Wavelet decomposition is performed on each intrinsic mode component to obtain multiple wavelet layers in each intrinsic mode component.

3. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 1, characterized in that, The calculation of the first noise level for each wavelet layer, based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer, includes: The energy concentration of each wavelet layer is obtained based on the energy deviation between the peak value of each wavelet layer and its neighboring detail coefficients; the neighboring detail coefficients of each detail coefficient peak indicate the detail coefficients within a preset time window before and after the occurrence of the detail coefficient peak. The degree of drastic change in the detail coefficients of each wavelet layer is obtained by considering the differences between the peak values ​​of adjacent detail coefficients in each wavelet layer. The energy concentration of each wavelet layer is fused with the degree of drastic change in its detail coefficients to obtain the first noise level of each wavelet layer.

4. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 3, characterized in that, The energy concentration of each wavelet layer is obtained based on the energy deviation between the peak value of each wavelet layer and its neighboring detail coefficients, including: Calculate the ratio of the energy corresponding to the peak value of each detail coefficient in each wavelet layer to the mean of the energy corresponding to multiple detail coefficients in its neighborhood, and obtain the energy deviation corresponding to the peak value of each detail coefficient. The energy concentration of each wavelet layer is obtained by taking the average of the energy deviations corresponding to the peak values ​​of all detail coefficients in each wavelet layer.

5. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 3, characterized in that, The method of determining the degree of drastic change in the detail coefficients of each wavelet layer based on the difference between the peak values ​​of adjacent detail coefficients includes: Arrange all the detail coefficient peaks in each wavelet layer in chronological order to obtain the detail coefficient peak sequence of each wavelet layer; Calculate the absolute difference between each pair of adjacent detail coefficient peaks in the detail coefficient peak sequence of each wavelet layer to obtain multiple detail coefficient peak difference values; The mean of the peak differences of multiple detail coefficients in each wavelet layer is calculated to obtain the degree of drastic change of the detail coefficients in each wavelet layer.

6. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 1, characterized in that, The step of determining the second noise level of each wavelet layer based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers includes: Based on the amplitude of each detail coefficient peak in each wavelet layer, the detail coefficient peaks in the detail coefficient peak sequence of each wavelet layer are clustered to obtain the peak cluster and low peak cluster corresponding to each wavelet layer; the peak cluster represents the set of detail coefficient peaks with high amplitude, and the low peak cluster represents the set of detail coefficient peaks with low amplitude. Arrange the peak values ​​of detail coefficients in the peak clusters corresponding to each wavelet layer in chronological order to obtain the peak distribution sequence of each wavelet layer. The second noise level of each wavelet layer is calculated based on the distance between the peak distribution sequence of each wavelet layer and the peak distribution sequences of other wavelet layers under the same intrinsic mode component.

7. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 6, characterized in that, The calculation of the second noise level for each wavelet layer based on the distance between the peak distribution sequence of each wavelet layer and the peak distribution sequences of other wavelet layers under the same intrinsic mode component includes: Calculate the Frescher distance between the peak point distribution sequence of each wavelet layer and the peak point distribution sequence of each other wavelet layer under the same intrinsic mode component to obtain multiple Frescher distances corresponding to each wavelet layer; The mean of multiple Fraser distances corresponding to each wavelet layer is calculated to obtain the second noise level of each wavelet layer.

8. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 1, characterized in that, The step of filtering each wavelet layer based on the fusion result of the first noise level and the second noise level, and reconstructing the microseismic signal based on all the filtered wavelet layers, includes: Calculate the high-frequency variability of each wavelet layer based on the variability of its detail coefficients. Calculate the mean of the high-frequency fluctuations of all wavelet layers in each intrinsic mode component to obtain the feature weights of each intrinsic mode component; The first noise level, the second noise level, and the feature weights of the intrinsic mode components corresponding to each wavelet layer are fused to obtain the comprehensive weight of each wavelet layer. Based on the comprehensive weight of each wavelet layer, the denoising threshold corresponding to each wavelet layer is adjusted to obtain the adjusted denoising threshold. Based on the adjusted denoising threshold, each wavelet layer is filtered to reconstruct the microseismic signal from all the filtered wavelet layers.

9. The hydrological model dynamic monitoring and simulation method based on multidimensional data as described in claim 8, characterized in that, The step of adjusting the denoising threshold corresponding to each wavelet layer based on the comprehensive weight of each wavelet layer to obtain the adjusted denoising threshold includes: The adjusted denoising threshold for each wavelet layer is obtained by multiplying the comprehensive weight of each wavelet layer with the preset reference wavelet threshold of its intrinsic mode component.

10. A dynamic monitoring and simulation system for hydrological models based on multidimensional data, characterized in that, The system includes: The decomposition module is used to decompose the microseismic signals in the mine to obtain multiple wavelet layers; The calculation module is used to calculate the first noise level of each wavelet layer based on the correlation between the peak distribution of detail coefficients and the degree of energy concentration in each wavelet layer; wherein, the first noise level indicates the degree of impulse interference to the wavelet layer, and the square of the detail coefficients of each wavelet layer is proportional to its corresponding energy; The determination module is used to determine the second noise level of each wavelet layer based on the similarity between the peak distribution of detail coefficients in each wavelet layer and the peak distribution of detail coefficients in other wavelet layers; the second noise level indicates the degree to which the wavelet layer is affected by high-frequency electromagnetic interference. The filtering module is used to filter each wavelet layer based on the fusion result of the first noise level and the second noise level, so as to reconstruct the microseismic signal based on all the filtered wavelet layers; the reconstructed microseismic signal is used for dynamic monitoring and simulation of mine water hazards.

Images