Device and method for automatically measuring artificial earthquake excitation time

Abstract

This application relates to the field of automatic measurement technology, specifically disclosing a device and method for automatically measuring the time of artificial earthquake excitation. The device utilizes a piezoelectric accelerometer to convert vibration signals into detection electrical signals, processes the detection electrical signals via a data acquisition card to obtain seismic waveform data, and then applies a GPS timestamp to the obtained seismic wave data. The device then uses a long- and short-time window algorithm to analyze the seismic waveform data to capture characteristic changes in the seismic waveform, thereby determining the start time of artificial earthquake excitation. This method avoids the phase shift problem inherent in the excitation response of a moving coil geophone, achieves high-precision automatic measurement of the time of artificial earthquake excitation, and is simple and convenient to use.

Description

Technical Field

[0001] The present application relates to the field of automatic measurement technology, and more specifically, to an automatic measurement device and method for artificial earthquake excitation timing. Background Art

[0002] Seismic imaging is an important method for detecting shallow surface structures. To achieve high-resolution shallow structure detection, artificial earthquakes are often used as the source. Artificial seismic excitation methods include explosive blasting, hammer impact on the ground, arc vaporization shock, and high-pressure air gun excitation. These artificial seismic excitation methods often require determining the excitation moment of the seismic waves, and high-precision excitation is crucial for seismic imaging. For shallow surface exploration (h < 100 m), high-frequency seismic imaging of at least 100 Hz is required, which requires millisecond-level accuracy of the seismic wave excitation moment. Furthermore, in engineering applications, the ease of determining the excitation moment affects the progress of exploration projects, necessitating the development of automated time measurement devices.

[0003] Currently, two methods are commonly used to determine the time of artificial earthquake excitation. The first method uses a moving-coil geophone-based earthquake excitation timing device. The moving-coil geophone converts the vibration signal into an electrical signal, which is then collected and compared with a threshold value to determine the time of earthquake excitation. This method is convenient because it does not require a circuit switch. However, the inherent hysteresis of the moving-coil response results in poor timing accuracy, typically within 3-5 milliseconds, making it difficult to meet the requirements of high-precision time measurement. The second method involves designing a circuit switch on the artificial earthquake excitation device, analogizing the closing of a switch when a heavy object strikes the ground. This method can accurately determine the time of artificial earthquake excitation. However, designing a circuit switch for artificial earthquake excitation devices using explosive blasting, heavy hammer impact, arc vaporization shock, and high-pressure air guns is difficult, making it inconvenient and impractical to operate. For artificial hammer impact methods that require a circuit switch, a constant voltage must be connected to the hammer head and an iron plate placed at the impact site to switch the circuit, which is inconvenient in practice.

[0004] Therefore, an optimized automatic measurement device and method for artificial earthquake excitation time are expected. Summary of the Invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an automatic measurement device and method for the artificial earthquake excitation moment, which uses a piezoelectric acceleration sensor to convert the vibration signal into a detection electrical signal, processes the detection electrical signal through a data acquisition card to obtain seismic waveform data, and then uses a long and short time window algorithm to analyze the seismic waveform data to capture the characteristic changes of the seismic waveform, thereby determining the starting time of the artificial earthquake excitation. In this way, the phase shift problem in the excitation response of the moving coil detector itself is avoided, and high-precision automatic measurement of the artificial earthquake excitation moment can be achieved, and the actual use is simple and convenient.

[0006] Accordingly, according to one aspect of the present application, a method for automatically measuring the artificial earthquake excitation time is provided, which includes:

[0007] Acquire a detection electrical signal collected by a piezoelectric acceleration sensor, wherein the time delay of the piezoelectric acceleration sensor is less than 3;

[0008] Processing the detection electrical signal using a data acquisition card to obtain seismic wave waveform data;

[0009] Signal processing is performed on the seismic wave waveform data to determine the starting time of artificial earthquake excitation.

[0010] According to another aspect of the present application, there is provided an automatic measurement device for artificial earthquake excitation time, comprising:

[0011] An electrical signal acquisition module, configured to acquire a detection electrical signal collected by a piezoelectric acceleration sensor, wherein the time delay of the piezoelectric acceleration sensor is less than 3;

[0012] an analog-to-digital conversion module, configured to process the detection electrical signal using a data acquisition card to obtain seismic wave waveform data;

[0013] The excitation starting time analysis module is used to perform signal processing on the seismic wave waveform data to determine the starting time of artificial earthquake excitation.

[0014] Compared to existing technologies, the device and method for automatically measuring the time of artificial earthquake excitation provided by this application utilizes a piezoelectric accelerometer to convert vibration signals into detection electrical signals. This detection electrical signal is then processed by a data acquisition card to obtain seismic waveform data. This seismic wave data is then timestamped with a GPS time stamp. Long- and short-time window algorithms are then used to analyze this seismic waveform data to capture characteristic changes in the seismic waveform, thereby determining the start time of artificial earthquake excitation. This method avoids the phase shift problem inherent in the excitation response of moving-coil geophones, enabling high-precision automatic measurement of the time of artificial earthquake excitation, while also being simple and convenient to use. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 Flowchart of the method for automatically measuring the artificial earthquake excitation time according to an embodiment of the present application.

[0017] Figure 2 This is a flowchart of step S3 in the method for automatically measuring the artificial earthquake excitation moment according to an embodiment of the present application.

[0018] Figure 3 Schematic diagram of data flow in step S3 of the method for automatic measurement of artificial earthquake excitation time according to an embodiment of the present application.

[0019] Figure 4 Flowchart of step S32 in the method for automatically measuring the artificial earthquake excitation moment according to an embodiment of the present application.

[0020] Figure 5 4 is a block diagram of an automatic measurement device for artificial earthquake excitation time according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0022] Example 1

[0023] As discussed in the background art above, the timing accuracy of earthquake excitation timing devices based on moving-coil geophones is limited. Specifically, a moving-coil geophone consists of a magnet, a coil, and a spring. Its operating principle is that when a seismic wave propagates to the moving-coil geophone, the coil, secured by the spring, undergoes mechanical movement (oscillation), disrupting the magnetic field generated by the magnet, generating an electrical signal. This mechanical movement of the coil takes time. Practical experience and analysis of the frequency response curves of moving-coil geophones show that moving-coil sensors exhibit varying phase delays for vibration signals of different frequencies, typically ranging from 3-5 milliseconds. Furthermore, this device determines the time of earthquake excitation by collecting electrical signals and performing threshold comparisons. Since the amplitude of seismic waves increases from weak to strong, the artificial earthquake is already triggered before the threshold is exceeded, resulting in a time delay. Furthermore, since a moving-coil geophone collects both noise and seismic waves generated by an artificial source, the comparison of the collected electrical signals with the threshold fails to distinguish between noise and seismic wave signals generated by the artificial earthquake, making it prone to mixing noise and seismic wave signals generated by the artificial earthquake.

[0024] To address the above technical issues, this application proposes an optimized method for automatically measuring the excitation moment of an artificial earthquake. This method uses a piezoelectric accelerometer to convert vibration signals into detection electrical signals. This detection electrical signal is processed by a data acquisition card to obtain seismic waveform data. The obtained seismic wave data is then timestamped with a GPS time stamp. A long- and short-time window algorithm is then used to analyze the seismic waveform data to capture characteristic changes in the seismic waveform, thereby determining the start time of the artificial earthquake excitation. This method avoids the phase shift problem inherent in the excitation response of a moving-coil geophone, enabling high-precision automatic measurement of the artificial earthquake excitation moment. Furthermore, the method is simple and convenient to use.

[0025] Figure 1 FIG. 1 is a flow chart of a method for automatically measuring the artificial earthquake excitation time according to an embodiment of the present application. Figure 1 As shown, the method for automatically measuring the artificial earthquake excitation moment according to the embodiment of the present application includes the following steps: S1, acquiring a detection electrical signal collected by a piezoelectric acceleration sensor, wherein the time delay of the piezoelectric acceleration sensor is less than 3; S2, processing the detection electrical signal using a data acquisition card to obtain seismic wave waveform data; S3, performing signal processing on the seismic wave waveform data to determine the starting moment of the artificial earthquake excitation.

[0026] In the above-mentioned method for automatically measuring the excitation moment of an artificial earthquake, step S1 obtains a detection electrical signal collected by a piezoelectric accelerometer, wherein the time delay of the piezoelectric accelerometer is less than 3. It should be understood that traditional moving coil geophones have the problem of excitation response phase shift, which limits measurement accuracy. However, piezoelectric accelerometers have advantages in time delay and signal conversion stability, and can provide high-quality original signals for subsequent processing. In a specific example of the present application, the piezoelectric material of the piezoelectric accelerometer is used to convert the measured vibration signal into the detection electrical signal through the piezoelectric effect. The time delay of the piezoelectric accelerometer is less than 3, and the frequency response range is 0.5-2500 Hz. Specifically, the piezoelectric material inside the piezoelectric accelerometer has a unique piezoelectric effect. When the sensor detects vibrations generated by an artificial earthquake, the piezoelectric material deforms due to mechanical stress. This deformation causes the charge distribution inside the piezoelectric material to change, thereby generating a detection electrical signal proportional to the vibration intensity. Taking the common quartz crystal piezoelectric material as an example, when subjected to external force, the centers of positive and negative charges in its crystal structure shift relative to each other, generating a charge on the material's surface. This charge is closely related to the magnitude and direction of the applied force. This direct conversion of mechanical energy into electrical energy not only offers a fast response but also improves the stability of signal conversion, facilitating rapid signal acquisition and processing.

[0027] In the above-mentioned method for automatically measuring the time of artificial earthquake excitation, step S2 involves processing the detection electrical signal using a data acquisition card to obtain seismic waveform data. It should be understood that the raw detection electrical signal output by the piezoelectric accelerometer is typically a continuous analog signal and contains a large amount of noise and interference information, making it relatively chaotic. Therefore, to convert it into intuitive, analyzable seismic waveform data, the present application further utilizes a data acquisition card to process the detection electrical signal. It should be understood that a data acquisition card is a device that converts analog signals into digital signals and processes and stores the digital signals. In a specific example of the present application, the data acquisition card comprises an ADC, a GPS module, an MCU main control system, and a field-programmable gate array (FPGA). The FPGA is used to implement data caching and control circuitry, while the MCU main control system controls the data acquisition card, drives the ADC (analog-to-digital converter) to acquire electrical signals, converts the continuous analog detection electrical signal output by the piezoelectric accelerometer into a discrete digital signal, and uses a built-in GPS (Global Positioning System) module to stamp the detection electrical signal with a GPS timestamp to obtain the seismic waveform data. Here, the GPS module can obtain accurate time information by receiving satellite signals, and then attach the time information to the converted digital signal in the form of a timestamp to obtain seismic wave waveform data with accurate time marks, providing a reliable time reference for subsequent seismic wave waveform data analysis.

[0028] In the above-mentioned method for automatically measuring the time of artificial earthquake excitation, step S3 involves signal processing the seismic waveform data to determine the start time of artificial earthquake excitation. In a specific example of the present application, a long-short time window algorithm is used to analyze the seismic waveform data to determine the start time of artificial earthquake excitation. It should be understood that before artificial earthquake excitation, the seismic wave signal is primarily background noise, with relatively stable characteristics and minimal energy and amplitude changes. However, at the time of excitation, the seismic wave signal undergoes significant changes, with a sharp increase in energy and amplitude. The long-short time window algorithm performs sliding calculations on the seismic wave waveform data by setting two time windows of different lengths (a short window and a long window). The short window is relatively short, typically tens of milliseconds, and is used to quickly capture instantaneous signal changes. The long window is longer, typically between hundreds of milliseconds and several seconds, and is used to reflect the long-term average characteristics of the signal. By comparing the energy or other characteristic parameters of the signals within the two windows, such as amplitude and frequency, the start time of artificial earthquake excitation can be determined when the signal characteristics within the short window show a significant change relative to the long window. For example, when the energy value within the short-time window exceeds a certain multiple of the energy value within the long-time window, and the duration reaches a certain threshold, an artificial earthquake can be considered to have occurred. By using a long- and short-time window algorithm to analyze seismic waveform data, this application can accurately and reliably identify the onset of artificial earthquake excitation from complex seismic signals. This avoids measurement errors caused by the phase shift of the moving coil detector's own excitation response, and achieves high-precision automatic measurement of the artificial earthquake excitation moment. Furthermore, the entire measurement process is automatically completed by a computer program, eliminating the need for human intervention, greatly improving measurement efficiency and accuracy.

[0029] Example 2

[0030] In particular, considering that the traditional long-short time window algorithm mainly performs basic statistical analysis on the waveform data in the short-time window and the long-time window, such as calculating energy, amplitude, etc., it lacks a detailed description of the changes in waveform characteristics, and when comparing the long-short time window characteristics, it often adopts a relatively simple method, such as directly calculating the energy ratio of the long-short time window, etc., but lacks an in-depth exploration of the complex interactive relationship between the long-short time window characteristics. When encountering large noise interference or complex and changeable seismic wave waveforms, this simple comparative analysis method may not be able to accurately distinguish between effective signals and noise, resulting in deviations in the judgment of the starting time of artificial earthquake excitation. In this regard, the present application proposes an improved long-short time window feature comparative analysis algorithm, which uses a convolutional neural network to perform deep learning on the waveform features in the long and short time windows to capture the deeper waveform semantic change features in the short-time window and the long-time window, and through a semantic comparative analysis of the two based on core information anchoring, it realizes the judgment of the starting time of artificial earthquake excitation, thereby improving the accuracy and anti-interference ability of the measurement of the starting time of artificial earthquake excitation.

[0031] Figure 2 This is a flowchart of step S3 in the method for automatically measuring the artificial earthquake excitation moment according to an embodiment of the present application. Figure 3 FIG. 1 is a data flow diagram of step S3 in the method for automatically measuring the artificial earthquake excitation time according to an embodiment of the present application. Figure 2 and Figure 3 As shown, the step S3 includes: S31, performing multi-scale waveform feature extraction on the seismic wave waveform data to obtain a seismic wave short-time window waveform semantic coding feature vector and a seismic wavelength time window waveform semantic coding feature vector; S32, performing a long-short time window waveform semantic comparison analysis based on core information anchoring on the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector to obtain a seismic wavelength-short time window waveform semantic comparison coding vector; S33, inputting the seismic wavelength-short time window waveform semantic comparison coding vector into a decoder-based excitation starting time estimation module to obtain the artificial earthquake excitation starting time.

[0032] Specifically, the step S31 performs multi-scale waveform feature extraction on the seismic wave waveform data to obtain a seismic wave short-time window waveform semantic coding feature vector and a seismic wavelength time window waveform semantic coding feature vector. It should be understood that since the traditional long and short time window algorithms mainly perform basic statistical analysis on the waveform data within the short time window and the long time window, there is a lack of detailed description of the waveform features. To this end, the present application introduces a deep learning algorithm to mine the waveform features within the long and short time windows of the seismic wave waveform data. In a specific example of the present application, the seismic wave waveform data is subjected to short-time window waveform feature extraction based on one-dimensional convolution coding to obtain a seismic wave short-time window waveform semantic coding feature vector; the seismic wave waveform data is subjected to long-time window waveform feature extraction based on one-dimensional convolution coding to obtain a seismic wavelength time window waveform semantic coding feature vector. Those skilled in the art should know that one-dimensional convolutional coding is the application of convolutional neural networks (CNNs) on one-dimensional data. It uses a one-dimensional convolution kernel to slide along the time direction on the seismic wave waveform data, and performs convolution operations on the waveform data at different times to learn local patterns in the data, such as the rising edge and falling edge of the waveform, and enhances the nonlinear expression of the features through the processing of activation functions, thereby mapping the original waveform data into a vector representation in a high-dimensional semantic feature space. In this application, by using one-dimensional convolution kernels of short-time windows and long-time windows to perform convolution coding on the seismic wave waveform data respectively, the different feature representations of the seismic wave waveform data in the short-time window and the long-time window can be learned respectively, and the corresponding seismic wave short-time window waveform semantic coding feature vector and seismic long-time window waveform semantic coding feature vector are obtained. Among them, the semantic coding feature vector of the seismic wave short-time window waveform mainly reflects the instantaneous change characteristics of the seismic wave signal, such as the waveform's mutation point, high-frequency components, etc.; while the semantic coding feature vector of the seismic wave long-time window waveform focuses more on reflecting the long-term average characteristics and overall trend of the seismic wave signal, such as the waveform's stable state, low-frequency components, etc.

[0033] Specifically, step S32 performs a long-short time window waveform semantic comparison analysis based on core information anchoring on the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector to obtain a seismic wavelength-short time window waveform semantic comparison coding vector. It should be understood that the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector respectively provide the instantaneous dynamic characteristics of the seismic wave signal and the long-term background information of the seismic wave signal. By comparing and analyzing the two, it is possible to capture the subtle changes in the seismic wave signal before and after the artificial earthquake excitation, thereby achieving accurate judgment of the starting time of the artificial earthquake excitation. In order to improve the accuracy and efficiency of comparative analysis, this application proposes a long-short time window waveform semantic comparative analysis method based on core information anchoring, which identifies the key information components of the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector by performing core information mining on the seismic wave short-time window waveform semantic coding feature vector, and captures the deep-level correlation and difference between the short-time window and long-time window waveform features through multi-granularity feature comparison interaction to generate a seismic wavelength-short-time window waveform semantic comparative coding vector.

[0034] Figure 4 FIG. 1 is a flow chart of step S32 in the method for automatically measuring the artificial earthquake excitation time according to an embodiment of the present application. Figure 4 As shown, the step S32 includes: S321, respectively extracting the core information of the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector to obtain the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector; S322, performing feature granularity response interactive coding on the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector to obtain the seismic wavelength-short-time window waveform semantic feature granularity response interactive coding vector; S323, performing feature value granularity response interactive coding on the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector to obtain the seismic wavelength-short-time window waveform semantic feature value granularity response interactive coding vector; S324, fusing the seismic wavelength-short-time window waveform semantic feature granularity response interactive coding vector and the seismic wavelength-short-time window waveform semantic feature value granularity response interactive coding vector to obtain the seismic wavelength-short-time window waveform semantic comparison coding vector.

[0035] In a specific example of the present application, step S321 includes: first, constructing a semantic autocorrelation association matrix of the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector to obtain a seismic wave short-time window waveform semantic autocorrelation association matrix and a seismic wavelength time window waveform semantic autocorrelation association matrix, which is expressed as follows:

[0036]

[0037]

[0038] in, represents the transpose of a vector, Represents the semantic encoding feature vector of the seismic wave short-time window waveform, Represents the semantic encoding feature vector of the seismic wavelength time window waveform, represents the semantic autocorrelation matrix of the seismic wave short-time window waveform, represents the semantic autocorrelation matrix of the seismic wavelength time window waveform, is a linear mapping function.

[0039] Here, by constructing the semantic autocorrelation matrix of the semantic encoding feature vector of the seismic wave short-time window waveform and the semantic encoding feature vector of the seismic wavelength time window waveform, the internal correlation structure of the two is revealed, so as to better understand the internal logic of the change of seismic waves over time. In a specific implementation, the mutual correlation between the component feature parts in the vector can be made explicit through the vector outer product operation to generate the semantic autocorrelation matrix of the seismic wave short-time window waveform and the semantic autocorrelation matrix of the seismic wavelength time window waveform, thereby capturing its internal semantic correlation pattern while retaining the global temporal structure of the seismic wave waveform characteristics.

[0040] Next, the seismic wave short-time window waveform semantic autocorrelation association matrix and the seismic wavelength time window waveform semantic autocorrelation association matrix are respectively input into the core information anchoring network based on autocorrelation decoupling to obtain the seismic wave short-time window waveform semantic core information anchoring coding vector and the seismic wavelength time window waveform semantic core information anchoring coding vector, which are expressed as follows:

[0041]

[0042]

[0043] in, Indicates the core information anchoring network, represents the decoupling function, 、 、 and They represent the first, second, and third elements in the semantic autocorrelation matrix of the seismic wave short-time window waveform. and row vectors, is the number of rows of the semantic autocorrelation matrix of the seismic wave short-time window waveform, 、 、 and They represent the first, second, and third elements in the semantic autocorrelation matrix of the seismic wavelength time window waveform. and row vectors, and They represent the semantic weight matrix of seismic wave short-time window waveform and the semantic weight matrix of seismic wave long-time window waveform respectively, and They represent the semantic bias vector of seismic wave short-time window waveform and the semantic bias vector of seismic wave long-time window waveform respectively, represents matrix multiplication, express The corresponding seismic wave short-time window waveform semantic importance scoring factor, express The corresponding seismic wavelength window waveform semantic importance scoring factor, represents the seismic wave short-time window local waveform feature importance score conversion vector, represents the seismic wavelength window local waveform feature importance score conversion vector, express The corresponding normalized seismic wave short-time window waveform semantic importance scoring factor, express The corresponding normalized seismic wavelength time window waveform semantic importance scoring factor, represents the sigmoid function, Represents the anchor coding vector of the semantic core information of the seismic wave short-time window waveform, Represents the anchor coding vector of the semantic core information of the seismic wavelength time window waveform.

[0044] That is, in order to extract the key core information components in the semantic features of the seismic wave short-time window waveform and the semantic features of the seismic wavelength time window waveform, the present application first performs fine-grained feature decoupling on the seismic wave short-time window waveform semantic autocorrelation matrix and the seismic wavelength time window waveform semantic autocorrelation matrix to achieve feature segmentation, and extracts and purifies the internal feature information through local feature importance evaluation and feature weighted aggregation, thereby highlighting the core feature information in the seismic wave short-time window waveform semantic autocorrelation matrix and the seismic wavelength time window waveform semantic autocorrelation matrix, and suppressing the interference of non-redundant information, thereby realizing the temporal core feature anchoring of the seismic wave short-time window waveform semantic features and the long-time window waveform semantic features, and obtaining the seismic wave short-time window waveform semantic core information anchoring coding vector and the seismic wavelength time window waveform semantic core information anchoring coding vector.

[0045] Then, the deep-level interaction pattern between the semantic features of the short-time window waveform of the seismic wave and the semantic features of the long-time window waveform is captured by further interactive response modeling analysis of the semantic core information anchoring coding vector of the seismic wave short-time window waveform and the semantic core information anchoring coding vector of the seismic wave long-time window waveform. Here, the application adopts a multi-level interactive response analysis method, which reveals the interactive influence of the semantic features of the short-time window waveform of the seismic wave and the semantic features of the long-time window waveform at different scales by conducting in-depth analysis at the two levels of feature granularity and feature value granularity, and more accurately models the dynamic change process of the seismic wave signal before and after artificial earthquake excitation.

[0046] In a specific example of the present application, step S322 is expressed as follows:

[0047]

[0048] in, express function, represents the weight matrix of seismic wavelength-short-time window waveform semantic interaction features, represents the seismic wavelength-short time window waveform semantic interaction feature bias vector, Represents the interaction encoding vector of seismic wavelength-short-time window waveform semantic feature granularity response.

[0049] In a specific example of the present application, step S323 is expressed as follows:

[0050]

[0051] in, Represents the interaction coding vector of seismic wavelength-short-time window waveform semantic feature value granularity response.

[0052] Here, at the feature granularity level, the present application constructs a feature interaction layer based on a neural network architecture, and captures the temporal correlation changes and interaction patterns of the two in the feature space by learning the feature combination relationship between the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector, and generates a seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector. At the feature value granularity level, the temporal interaction modeling between the seismic wave short-time window waveform semantic features and the long-time window waveform semantic features returns to a more basic numerical operation, and the subtle correlation and mutual dependence between the two in temporal changes are characterized by performing element-by-element interactive response coding on the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector, and generating a seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector.

[0053] In a specific example of the present application, step S324 includes: cascading and fusing the seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector and the seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector to obtain the seismic wavelength-short-time window waveform semantic contrast coding vector, which is expressed as follows:

[0054]

[0055] in, Indicates cascade, Represents the semantic contrast encoding vector of seismic wavelength-short time window waveform.

[0056] Specifically, by cascading and fusing the seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector and the seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector, the interaction response information at different levels is comprehensively considered to generate a seismic wavelength-short-time window waveform semantic comparison coding vector. This approach fully utilizes the waveform feature information of seismic wave signals at different time scales, focusing on the waveform feature changes that are most critical for determining the start time of artificial seismic excitation, effectively eliminating noise interference, and thus improving the accuracy and robustness of artificial seismic excitation start time analysis.

[0057] In particular, in a preferred example of the present application, the step S324 includes: first, performing a feature distribution gradient constraint correction based on bidirectional interaction on the seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector and the seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector to obtain an optimized seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector and an optimized seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector, which can be expressed as follows:

[0058]

[0059]

[0060] in, yes Middle The eigenvalues ​​at the positions, yes Middle The eigenvalues ​​at the positions, represents the cosine function, yes The corresponding optimized eigenvalue is the optimized seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector. The eigenvalues ​​at the positions, yes The corresponding optimized eigenvalue is the optimized seismic wavelength-short-time window waveform semantic eigenvalue granularity response interaction coding vector. The eigenvalues ​​at each position.

[0061] Then, the optimized seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector and the optimized seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector are cascaded and fused to obtain the seismic wavelength-short-time window waveform semantic comparison coding vector.

[0062] In particular, here, considering the difference between the modeling representations of feature granularity interaction and eigenvalue granularity interaction may cause unstable perturbations in the feature manifold interface of the fused seismic wavelength-short-time window waveform semantic contrast coding vector with multi-granularity information expression. Therefore, the present application further uses the seismic wavelength-short-time window waveform semantic eigenvalue granularity response interaction coding vector as an extensibility constraint representation, so as to model the interface shape perturbation deviation of the overall distribution of feature granularity interaction based on the growth exponent representation under the extensibility constraint for the overall feature interaction distribution growth under the eigenvalue diffusion process. That is, the interaction interface gradient under the eigenvalue granularity is used as the perturbation contribution factor, and the growth exponent stabilization contribution of the gradient diffusion under the eigenvalue mutual dependence is determined to achieve the growth mode gradient correction under the dominance of perturbation stabilization, thereby improving the manifold interface stability of the seismic wavelength-short-time window waveform semantic contrast coding vector.

[0063] Specifically, step S33 inputs the seismic wavelength-short-time window waveform semantic comparison coding vector into the decoder-based excitation start time estimation module to obtain the artificial earthquake excitation start time. Specifically, the decoder is based on a neural network structure. After receiving the seismic wavelength-short-time window waveform semantic comparison coding vector as input, it performs feature decoding on the seismic wavelength-short-time window waveform semantic comparison coding vector through a series of fully connected layers and activation functions to learn the dynamic waveform changes of the seismic wave signal before and after artificial earthquake excitation, so as to gradually restore the artificial earthquake excitation start time information contained in the seismic wave signal, thereby outputting the decoded value of the artificial earthquake excitation start time. Here, during the training process, the decoder uses a large amount of sample data with known excitation start time to optimize its network parameters by minimizing the error between the predicted start time and the actual start time, so as to accurately estimate the excitation start time of unknown seismic data. This method not only improves the automation level of artificial earthquake excitation start time analysis, but also significantly enhances the accuracy and efficiency of the analysis.

[0064] In summary, the method for automatically measuring the time of artificial earthquake excitation according to the embodiments of the present application is described. It utilizes a piezoelectric accelerometer to convert vibration signals into detection electrical signals, processes the detection electrical signals via a data acquisition card to obtain seismic waveform data, and then applies a GPS timestamp to the obtained seismic wave data. The seismic waveform data is then analyzed using a long- and short-time window algorithm to capture characteristic changes in the seismic waveform, thereby determining the start time of artificial earthquake excitation. This method avoids the phase shift problem inherent in the excitation response of moving-coil geophones, enables high-precision automatic measurement of the time of artificial earthquake excitation, and is simple and convenient to use.

[0065] Furthermore, the present application also provides an automatic measurement device for artificial earthquake excitation time.

[0066] Figure 5 FIG. 1 is a block diagram of an automatic measurement device for artificial earthquake excitation time according to an embodiment of the present application. Figure 5 As shown, the automatic measurement device 100 for the artificial earthquake excitation time according to the embodiment of the present application includes: an electrical signal acquisition module 110, which is used to obtain the detection electrical signal collected by the piezoelectric acceleration sensor, and the time delay of the piezoelectric acceleration sensor is less than 3; an analog-to-digital conversion module 120, which is used to use a data acquisition card to process the detection electrical signal to obtain seismic wave waveform data; and an excitation starting time analysis module 130, which is used to perform signal processing on the seismic wave waveform data to determine the starting time of artificial earthquake excitation.

[0067] Here, those skilled in the art will appreciate that the specific operations of each module in the above-mentioned automatic measurement device for artificial earthquake excitation time have been described in detail in the description of the above-mentioned automatic measurement method for artificial earthquake excitation time, and therefore, repeated description thereof will be omitted.

[0068] Comparative Example 1

[0069] As to the effectiveness of this device, the artificial earthquake excitation moment obtained by the high-precision automatic measurement device for artificial earthquake excitation moment described in this application is compared with the moment obtained by designing a circuit switching method. The circuit switching method is composed of an iron hammer, an iron plate and a GNSS taming clock. The GNSS taming clock can timestamp the external input pulse signal with a timestamp accuracy of 10ns. The iron hammer, the iron plate and the GNSS taming clock are connected to the same ground, a 3.3V voltage is connected to the iron hammer, and the iron plate is connected to the external input pulse terminal of the GNSS taming clock. When the iron hammer hits the iron plate, an external pulse is formed, and the GNSS taming clock timestamps the pulse. The electrical signal pulse has low latency, and the error of the circuit switching method is within 15ns. The timestamp can be considered to be an accurate moment. Table 1 shows the comparison results of the moments obtained by the high-precision automatic measurement device for artificial earthquake excitation moment described in this application and the circuit switching method.

[0070] Table 1 Comparison of automatic measurement results and accurate time of the device of the present invention

[0071]

[0072] It should be understood that the artificial earthquake excitation time obtained by the designed circuit switching method has a high degree of accuracy. As can be seen from Table 1, the artificial earthquake excitation time obtained by the high-precision automatic measurement device for artificial earthquake excitation time described in this application is within 1 millisecond of the time obtained by the designed circuit switching method, demonstrating the high accuracy of the device of the present invention. In addition, the device of the present invention uses a non-contact automated measurement method, which is simpler and safer to operate than the traditional circuit switching method, and is practically feasible and convenient.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for automatically measuring the time of artificial earthquake excitation, characterized in that: include: Acquire a detection electrical signal collected by a piezoelectric acceleration sensor, wherein the time delay of the piezoelectric acceleration sensor is less than 3; Processing the detection electrical signal using a data acquisition card to obtain seismic wave waveform data; Performing signal processing on the seismic waveform data to determine the start time of artificial earthquake excitation; Acquiring the detection electrical signal collected by the piezoelectric acceleration sensor, including: The piezoelectric material of the piezoelectric acceleration sensor is used to convert the measured vibration signal into the detection electrical signal through the piezoelectric effect; Processing the detection electrical signal using a data acquisition card to obtain seismic wave waveform data, including: adding a GPS time stamp to the detection electrical signal to obtain the seismic wave waveform data; Signal processing is performed on the seismic waveform data to determine the start time of artificial earthquake excitation, including: Performing multi-scale waveform feature extraction on the seismic wave waveform data to obtain a seismic wave short-time window waveform semantic coding feature vector and a seismic wave long-time window waveform semantic coding feature vector; Performing a long-short time window waveform semantic comparison analysis based on core information anchoring on the seismic wave short time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector to obtain a seismic wavelength-short time window waveform semantic comparison coding vector; The seismic wavelength-short time window waveform semantic comparison coding vector is input into the decoder-based excitation starting time estimation module to obtain the artificial earthquake excitation starting time.

2. The method for automatically measuring the artificial earthquake excitation time according to claim 1, characterized in that: Signal processing is performed on the seismic waveform data to determine the starting time of artificial earthquake excitation, including: using a long-short time window algorithm to analyze the seismic waveform data to determine the starting time of artificial earthquake excitation.

3. The method for automatically measuring the artificial earthquake excitation time according to claim 2, characterized in that: Performing multi-scale waveform feature extraction on the seismic wave waveform data to obtain a seismic wave short-time window waveform semantic coding feature vector and a seismic wave long-time window waveform semantic coding feature vector, including: Performing short-time window waveform feature extraction based on one-dimensional convolution coding on the seismic wave waveform data to obtain the seismic wave short-time window waveform semantic coding feature vector; The seismic wave waveform data is subjected to long-time window waveform feature extraction based on one-dimensional convolution coding to obtain the seismic wave long-time window waveform semantic coding feature vector.

4. The method for automatically measuring the artificial earthquake excitation time according to claim 3, characterized in that: Performing a long-short time window waveform semantic comparison analysis based on core information anchoring on the seismic wave short time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector to obtain a seismic wavelength-short time window waveform semantic comparison coding vector, including: Extracting the core information of the seismic wave short-time window waveform semantic coding feature vector and the seismic wave wavelength time window waveform semantic coding feature vector respectively to obtain the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wave wavelength time window waveform semantic core information anchor coding vector; Performing feature granularity response interactive coding on the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector to obtain a seismic wavelength-short-time window waveform semantic feature granularity response interactive coding vector; Performing eigenvalue granularity response interactive coding on the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector to obtain a seismic wavelength-short-time window waveform semantic eigenvalue granularity response interactive coding vector; The seismic wavelength-short time window waveform semantic feature granularity response interactive coding vector and the seismic wavelength-short time window waveform semantic feature value granularity response interactive coding vector are fused to obtain the seismic wavelength-short time window waveform semantic comparison coding vector.

5. The method for automatically measuring the artificial earthquake excitation time according to claim 4, characterized in that: Extracting the core information of the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector respectively to obtain the seismic wave short-time window waveform semantic core information anchor coding vector and the seismic wavelength time window waveform semantic core information anchor coding vector, including: Constructing a semantic autocorrelation association matrix of the seismic wave short-time window waveform semantic coding feature vector and the seismic wavelength time window waveform semantic coding feature vector to obtain a seismic wave short-time window waveform semantic autocorrelation association matrix and a seismic wavelength time window waveform semantic autocorrelation association matrix; The seismic wave short-time window waveform semantic autocorrelation association matrix and the seismic wavelength time window waveform semantic autocorrelation association matrix are respectively input into the core information anchoring network based on autocorrelation decoupling to obtain the seismic wave short-time window waveform semantic core information anchoring coding vector and the seismic wavelength time window waveform semantic core information anchoring coding vector.

6. The method for automatically measuring the artificial earthquake excitation time according to claim 5, characterized in that: The seismic wavelength-short time window waveform semantic feature granularity response interaction coding vector and the seismic wavelength-short time window waveform semantic feature value granularity response interaction coding vector are integrated to obtain the seismic wavelength-short time window waveform semantic comparison coding vector, including: Performing bidirectional interactive feature distribution gradient constraint correction on the seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector and the seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector to obtain an optimized seismic wavelength-short-time window waveform semantic feature granularity response interaction coding vector and an optimized seismic wavelength-short-time window waveform semantic feature value granularity response interaction coding vector; The optimized seismic wavelength-short-time window waveform semantic feature granularity response interactive coding vector and the optimized seismic wavelength-short-time window waveform semantic feature value granularity response interactive coding vector are cascaded and fused to obtain the seismic wavelength-short-time window waveform semantic comparison coding vector.

7. An automatic measuring device for artificial earthquake excitation time, used to perform the method according to any one of claims 1 to 6, characterized in that: include: An electrical signal acquisition module, configured to acquire a detection electrical signal collected by a piezoelectric acceleration sensor, wherein the time delay of the piezoelectric acceleration sensor is less than 3; an analog-to-digital conversion module, configured to process the detection electrical signal using a data acquisition card to obtain seismic wave waveform data; The excitation starting time analysis module is used to perform signal processing on the seismic wave waveform data to determine the starting time of artificial earthquake excitation.

Images