Earthquake geomagnetic observation data acquisition and transmission method and system

Abstract

The invention discloses an earthquake geomagnetic observation data acquisition and transmission method and system, and relates to the technical field of earthquake monitoring, and the method comprises the steps: collecting original waveform data with an event mark, carrying out the cooperative pulse distribution based on a pulse neural network calculation environment, and obtaining an enhanced pulse time sequence; a pulse filtering algorithm based on a dynamic amplitude threshold is adopted, the information density and the time sequence consistency of the enhanced pulse time sequence are optimized in combination with a phase-locked loop synchronization method, and an optimized pulse sequence with a high signal-to-noise ratio is generated; and packaging the optimized pulse sequence into a standard transmission frame, transmitting the standard transmission frame to a seismic detection center, and performing decoding reconstruction through a CNN-LSTM model to generate high-fidelity seismic geomagnetic observation data. Through the CNN-LSTM model, the spatial features of the transmission frame are extracted by using the CNN convolutional layer, and the time sequence dependency relationship is learned in combination with the LSTM network, so that the precise reconstruction of the compressed transmission data is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of earthquake monitoring, in particular to a method and system for collecting and transmitting earthquake geomagnetic observation data. BACKGROUND

[0002] The collection and transmission of earthquake geomagnetic observation data is a key technical link in the earthquake monitoring and early warning system. Currently, the field generally uses geomagnetic sensor arrays based on fixed threshold triggering, combined with traditional ADC sampling and wireless transmission technology to achieve data collection. The conventional method continuously collects geomagnetic data at a fixed frequency (usually 50-200Hz) through a magnetoresistive sensor or a fluxgate magnetometer, triggers data recording when the magnetic field strength exceeds the preset threshold, and transmits it to the data center through LPWAN technologies such as LoRa or NB-IoT. In terms of data processing, wavelet transform combined with FIR filtering is commonly used for preprocessing, and data compression algorithms based on compressed sensing are used to improve transmission efficiency. This kind of technical scheme has formed a standardized application in the construction of seismic network, which can meet the basic needs of geomagnetic anomaly monitoring.

[0003] However, the existing method still has room for improvement in dealing with sudden geological events. On the one hand, the fixed threshold triggering mechanism is difficult to adapt to the dynamic changes of the magnetic field gradient in complex environments, which may lead to missed detection of weak precursor signals or false triggering of electromagnetic interference. On the other hand, the traditional data processing flow is not sufficient for feature extraction of pulse-type geomagnetic signals, and the attenuation of high-frequency components during transmission may affect the accuracy of subsequent analysis. Especially in a strong electromagnetic interference environment, the existing system needs to improve the ability to maintain the timing characteristics of the signal. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a method for collecting and transmitting earthquake geomagnetic observation data to solve the problem of accurate capture and high-fidelity transmission of geomagnetic anomaly events.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a method for collecting and transmitting earthquake geomagnetic observation data, which includes dynamically triggering a high-voltage pulse signal and generating an event marker data packet according to the rate of change of the environmental magnetic field gradient;

[0008] Collecting raw waveform data with event markers, calculating the coordinated pulse firing of the environment based on the pulse neural network, and obtaining an enhanced pulse timing sequence;

[0009] The pulse filtering algorithm based on dynamic amplitude threshold is combined with the phase-locked loop synchronization method to optimize the information density and timing consistency of the enhanced pulse timing sequence, and an optimized pulse sequence with high signal-to-noise ratio is generated.

[0010] The optimized pulse sequence is encapsulated into a standard transmission frame and transmitted to a seismic detection center, and meanwhile, the CNN-LSTM model is used for decoding and reconstruction to generate high-fidelity seismic geomagnetic observation data.

[0011] As a preferred scheme of the method for collecting and transmitting seismic geomagnetic observation data, the step of dynamically triggering a high-voltage pulse signal and generating an event marker data packet according to the environmental magnetic field gradient change rate is as follows,

[0012] The environmental magnetic field gradient change rate is continuously monitored by a Terfenol-D alloy sheet;

[0013] Based on statistical analysis of historical geomagnetic gradient data, an effective event trigger threshold is defined;

[0014] When the detected environmental magnetic field gradient change rate is greater than or equal to the effective event trigger threshold, the magnetostrictive effect trigger signal drives the Terfenol-D alloy sheet to deform, and a high-voltage pulse signal is output by mechanically coupling a PZT-5H piezoelectric ceramic;

[0015] After the high-voltage pulse signal is triggered, the UTC timestamp, the instantaneous magnetic field intensity change rate, and the background magnetic field baseline are synchronously acquired, and are encapsulated into an event marker data packet in TLV format through an SPI interface.

[0016] As a preferred scheme of the method for collecting and transmitting seismic geomagnetic observation data, the step of collecting raw waveform data with event markers is as follows,

[0017] The raw waveform data with event markers includes three-axis seismic waveform, three-axis geomagnetic waveform, UTC timestamp, instantaneous magnetic field intensity change rate, and background magnetic field baseline.

[0018] As a preferred scheme of the method for collecting and transmitting seismic geomagnetic observation data, the step of performing cooperative pulse firing based on a pulse neural network computing environment to obtain an enhanced pulse timing sequence is as follows,

[0019] An Intel Loihi 2 neuro-morphic chip is used to initialize a pulse neural network computing environment composed of LIF neurons;

[0020] Based on the pulse neural network computing environment, the initial pulse signal is generated by analyzing the initial waveform data with time mark through the LIF neuron, triggering the coordinated pulse firing when the dynamic event trigger threshold is exceeded;

[0021] The firing timing and spatial distribution of the initial pulse signal are analyzed by a three-level decomposition algorithm based on Daubechies4 wavelet, and after extracting the time-frequency feature coefficients, uniform quantization coding is performed to generate an enhanced pulse timing sequence.

[0022] As a preferred scheme of the method for collecting and transmitting seismic geomagnetic observation data, the pulse filtering algorithm based on the dynamic amplitude threshold is combined with the phase-locked loop synchronization method to optimize the information density and timing consistency of the enhanced pulse timing sequence, and a high signal-to-noise ratio optimized pulse sequence is generated, and the steps are as follows,

[0023] Based on the historical geomagnetic gradient monitoring data, the amplitude threshold is defined by the sliding window statistical analysis method;

[0024] The dynamic amplitude filtering algorithm is performed on the enhanced pulse timing sequence to eliminate redundant pulses below the amplitude threshold, and a preliminarily screened pulse sequence is generated;

[0025] The preliminarily screened pulse sequence is phase calibrated by the voltage-controlled oscillator of the digital phase-locked loop, and a timing-aligned pulse sequence is output;

[0026] The pulse interval compression algorithm is performed on the timing-aligned pulse sequence to improve the information density by adjusting the firing frequency, and a high-density pulse sequence is generated;

[0027] The dynamic time warping algorithm is used to optimize the timing consistency of the high-density pulse sequence, and a high signal-to-noise ratio optimized pulse sequence is generated.

[0028] As a preferred scheme of the method for collecting and transmitting seismic geomagnetic observation data, the optimized pulse sequence is encapsulated into a standard transmission frame, which means adding a protocol header to the optimized pulse sequence and encapsulating it in ASN.1 encoding format to generate a standard transmission frame.

[0029] As a preferred scheme of the method for collecting and transmitting seismic geomagnetic observation data, the decoding and reconstruction of the CNN-LSTM model is performed to generate high-fidelity seismic geomagnetic observation data, and the steps are as follows,

[0030] The standard transmission frame is transmitted to the seismic detection center through LPWAN, the local spatial features in the standard transmission frame are extracted through the CNN convolution layer of the CNN-LSTM model, and the long-term timing dependence between local spatial features is learned through the gating mechanism of the LSTM network;

[0031] Based on the long-term time sequence dependence relationship between local space features, the standard transmission frame is decoded and reconstructed, and the pulse time sequence inversion algorithm is used to recover the original waveform data with event markers.

[0032] In a second aspect, the application provides a seismic geomagnetic observation data acquisition and transmission system, comprising a data marker module for dynamically triggering a high-voltage pulse signal and generating an event marker data packet according to the environmental magnetic field gradient change rate;

[0033] A pulse synergistic enhancement module is used to acquire original waveform data with event markers, perform synergistic pulse firing based on a pulse neural network computing environment, and obtain an enhanced pulse time sequence;

[0034] A pulse optimization module is used to adopt a pulse filtering algorithm based on a dynamic amplitude threshold and a phase-locked loop synchronization method to optimize the information density and time sequence consistency of the enhanced pulse time sequence, and generate an optimized pulse sequence with high signal-to-noise ratio;

[0035] A seismic waveform reconstruction module is used to encapsulate the optimized pulse sequence into a standard transmission frame and transmit it to a seismic detection center, and simultaneously decode and reconstruct it through a CNN-LSTM model to generate high-fidelity seismic geomagnetic observation data.

[0036] In a third aspect, the application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the computer program, when executed by the processor, implements any step of the method for seismic geomagnetic observation data acquisition and transmission according to the first aspect of the application.

[0037] In a fourth aspect, the application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any step of the method for seismic geomagnetic observation data acquisition and transmission according to the first aspect of the application.

[0038] The application has the following advantages: through a pulse neural network computing environment, the biological characteristics of LIF neurons are used to realize pulse synergistic firing of original waveform data with event markers, effectively enhancing the information expression capability of the pulse time sequence; through a CNN-LSTM model, the spatial features of the transmission frame are extracted using a CNN convolution layer, and the time sequence dependence relationship is learned using an LSTM network, thereby realizing accurate reconstruction of compressed transmission data. The two steps work synergistically to form a complete technical chain from data acquisition to transmission reconstruction, which not only ensures the real-time response capability of geomagnetic anomaly events, but also ensures the integrity and accuracy of the observation data. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0040] Fig. 1 Flow chart for the method for collecting and transmitting seismic geomagnetic observation data.

[0041] Fig. 2 Schematic diagram of the system for collecting and transmitting seismic geomagnetic observation data.

[0042] Fig. 3 Flow chart for event marking and pulse synergistic enhancement. DETAILED DESCRIPTION

[0043] In order to make the above-mentioned objects, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings.

[0044] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0045] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0046] REFERENCE Figs. 1-3 For one embodiment of the present application, the embodiment provides a method for collecting and transmitting seismic geomagnetic observation data, comprising the following steps:

[0047] S1, dynamically triggering a high-voltage pulse signal and generating an event marking data packet according to the environmental magnetic field gradient change rate;

[0048] The environmental magnetic field gradient change rate is continuously monitored by a Terfenol-D alloy sheet at a sampling frequency of 100 Hz;

[0049] The environmental magnetic field gradient change refers to the difference rate of the magnetic field intensity between two adjacent points in space (unit: nT / m·s -1 ), which is used to reflect geological activities or electromagnetic interference events;

[0050] Based on statistical analysis of historical geomagnetic gradient data, the effective event trigger threshold is defined, usually in the range of 0.1-1.0 nT.

[0051] It should be noted that the effective event refers to the magnetic field gradient anomaly caused by the stress change of rock induced by earthquake precursor, the piezomagnetic effect caused by crustal microfracture, or the local magnetic field disturbance caused by magma activity, etc.

[0052] Electromagnetic interference events refer to magnetic field anomaly fluctuations caused by non-geological activity factors, such as power equipment operation-induced power frequency interference, lightning discharge-induced transient magnetic field change, or solar activity-induced geomagnetic disturbance, etc. These interferences can cause abnormal fluctuations in the environmental magnetic field gradient change rate monitored by the Terfenol-D alloy sheet, which may affect the accuracy of the effective event trigger threshold determination.

[0053] When the detected environmental magnetic field gradient change rate is greater than or equal to the effective event trigger threshold, the magnetostrictive effect trigger signal drives the Terfenol-D alloy sheet to produce deformation, and a high-voltage pulse signal is output through mechanical coupling excitation of the PZT-5H piezoelectric ceramic.

[0054] Further, when the detected environmental magnetic field gradient change rate is greater than or equal to the effective event trigger threshold, the Terfenol-D alloy sheet produces magnetostrictive effect under the action of the magnetic field, and the alloy sheet length undergoes microscopic deformation; the deformation is directly transmitted to the PZT-5H piezoelectric ceramic through the mechanical coupling structure, and the piezoelectric ceramic generates polarization charge due to mechanical stress; the polarization charge is converted into a high-voltage pulse signal output through a signal amplification circuit, realizing high-fidelity conversion from magnetic field change to electrical signal and ensuring reliable capture and transmission of geological activity signals.

[0055] After the high-voltage pulse signal is excited, the UTC timestamp, instantaneous magnetic field intensity change rate, and background magnetic field baseline are simultaneously acquired, and the event marker data packet is packaged in TLV format through the SPI interface.

[0056] Further, after the high-voltage pulse signal is excited, the synchronous data acquisition process is immediately started: the UTC timestamp accurate to microseconds is acquired through the GPS receiver, the instantaneous magnetic field intensity change rate is captured in real time by the three-axis magnetic force sensor, and the background magnetic field baseline is recorded as a reference benchmark. The acquired UTC timestamp, instantaneous magnetic field intensity change rate, and background magnetic field baseline are transmitted to the data processing unit through the SPI interface and structuredly packaged in TLV (Tag-Length-Value) format: the Tag field identifies the data type, the Length field records the data length, and the Value field stores the actual numerical value. The packaged event marker data packet contains complete time information, magnetic field change parameters, and environmental reference data, forming a standardized data storage unit. The data packet structure ensures clear boundaries between fields, facilitating subsequent parsing and processing.

[0057] S2, collect original waveform data with event markers, perform coordinated spike firing based on a spiking neural network computing environment, and obtain an enhanced pulse timing sequence;

[0058] The original waveform data is synchronously collected by the triaxial seismic sensor and the triaxial geomagnetic sensor, and an event marker data packet is associated in real time during the collection process to generate original waveform data with event markers.

[0059] Further, the triaxial seismic sensor and the triaxial geomagnetic sensor are synchronously started for data collection at a sampling frequency of 100 Hz, the triaxial seismic sensor records seismic waveforms in X / Y / Z axial directions, and the triaxial geomagnetic sensor synchronously collects magnetic field waveforms in X / Y / Z axial directions; during the collection process, event marker data packets triggered by the Terfenol-D alloy sheet are received in real time, and the seismic waveform data, the geomagnetic waveform data, the UTC time stamp in the event marker data packet, the instantaneous magnetic field intensity change rate, and the background magnetic field baseline are time-aligned and parameter-associated to generate original waveform data with event markers containing triaxial seismic waveforms, triaxial geomagnetic waveforms, UTC time stamps, instantaneous magnetic field intensity change rates, and background magnetic field baselines.

[0060] The original waveform data with event markers includes triaxial seismic waveforms, triaxial geomagnetic waveforms, UTC time stamps, instantaneous magnetic field intensity change rates, and background magnetic field baselines.

[0061] An Intel Loihi 2 neuromorphic chip is used to initialize a spiking neural network computing environment composed of 128 LIF neurons.

[0062] Further, the specific implementation process of initializing the 128 LIF neuron computing environment of the Intel Loihi 2 neuromorphic chip is as follows: when the chip configures the neuron network through the NxCore instruction set, first, a two-dimensional grid topology is established to form a local connection cluster of physically adjacent neurons; on this basis, the basic parameters such as resting potential, firing threshold, and membrane time constant are set for each LIF neuron to ensure that the neurons have biologically reasonable dynamic response characteristics; then the STDP rule is used to initialize the synaptic connection weight, where the excitatory synapse promotes signal propagation and the inhibitory synapse maintains network balance; the asynchronous spike event routing engine optimizes the pulse transmission path based on the priority queue, and cooperates with the neuron parameter configuration to jointly build a low-delay communication network; the x86 coprocessor monitors the network state in real time through a dedicated bus, dynamically adjusts the global clock frequency according to the neuron firing frequency and energy consumption data, and forms a closed-loop control mechanism. The above process from physical connection, parameter configuration to dynamic regulation finally builds a spiking neural network computing environment with biological plasticity, realizing efficient spatiotemporal pattern processing capability.

[0063] It should be noted that a plasticity learning mechanism based on pulse timing dependence, for example, when the presynaptic neuron pulse arrives earlier than the postsynaptic neuron pulse, the synaptic weight is enhanced; otherwise, the weight is weakened.

[0064] Based on the pulse neural network computing environment, the 28 LIF neurons analyze the time-tagged initial waveform data, trigger the coordinated pulse firing when the dynamic event trigger threshold is exceeded, and generate the original pulse signal;

[0065] Further, the Intel Loihi 2 neuromorphic chip receives time-tagged initial waveform data input, and 28 LIF neurons process data synchronously in a preset two-dimensional grid topology; each LIF neuron continuously integrates the input pulse, and immediately fires a pulse when the membrane potential exceeds the dynamic event trigger threshold; the fired pulse propagates to adjacent neurons through the synaptic connection adjusted by the STDP rule, triggering a network cascade response; the neuron cluster that meets the coordinated firing condition synchronously outputs the original pulse signal through the asynchronous pulse event routing engine, and the signal time accuracy is strictly synchronized with the time tag of the original waveform data. This process realizes the pulse coding conversion of the initial waveform data, and provides an accurate spatiotemporal feature expression basis for the generation of enhanced pulse timing sequences.

[0066] It should be noted that the dynamic event trigger threshold is defined based on statistical analysis of historical geomagnetic gradient data, and the effective event trigger threshold range is usually 0.1-1.0nT.

[0067] The coordinated firing condition refers to a cluster response mechanism triggered when multiple LIF neurons in the pulse neural network reach the firing threshold within a certain time window (usually 5-10ms), for example, when analyzing earthquake precursor signals, more than 60% of the 28 LIF neurons fire pulses synchronously within an 8ms time window, which is determined as an effective coordinated firing event.

[0068] Through a three-level decomposition algorithm based on Daubechies4 wavelet, the firing timing and spatial distribution of the original pulse signal are analyzed at multiple scales, and after extracting the time-frequency feature coefficients, uniform quantization encoding is performed to generate enhanced pulse timing sequences.

[0069] Further, first, the firing timing of the original pulse signal is decomposed by three levels of wavelet, the first level decomposition extracts the 0-50Hz frequency band feature, the second level decomposition focuses on the 50-100Hz frequency band, and the third level decomposition analyzes the 100-200Hz frequency band; each level of decomposition extracts the spatial distribution characteristics of the pulse signal in X / Y / Z three axes, generating time-frequency feature coefficients containing time-frequency domain and spatial domain information (the role is to realize the multi-scale feature expression of the original pulse signal in the time-frequency domain); after normalization processing, the time-frequency feature coefficients are discretized to 0-255 integer values by 8bit uniform quantization coding; the quantized feature coefficients are sorted and reorganized according to the time stamp, and finally the enhanced pulse timing sequence with multi-scale time-frequency features is output. The sequence completely retains the space-time pattern characteristics of the original pulse signal, and enhances the analyzability of the high-frequency components.

[0070] It should be noted that the expression of the time-frequency feature coefficient is:

[0071]

[0072] Where C is the time-frequency feature coefficient, N is the total number of effective wavelet coefficients, j is the decomposition scale level, k is the translation factor, S is the original pulse signal, and ψ j,k represents the wavelet component of the Daubechies4 wavelet basis function at the jth level of decomposition scale and the kth translation position.

[0073] The effective wavelet coefficient refers to the detail coefficient that is not polluted by the boundary effect after Daubechies4 wavelet decomposition, and its number is determined by the signal length and the decomposition level.

[0074] It should be noted that the original pulse signal S is standardized by Z-score before participating in the calculation, in order to eliminate the sensor gain difference and ensure that the time-frequency feature coefficient C is dimensionless and the value is stable.

[0075] S3, using a pulse filtering algorithm based on dynamic amplitude threshold, combined with a phase-locked loop synchronization method to optimize the information density and timing consistency of the enhanced pulse timing sequence, generating an optimized pulse sequence with high signal-to-noise ratio.

[0076] Based on historical geomagnetic gradient monitoring data, the amplitude threshold is defined by sliding window statistical analysis method;

[0077] Further, the time window is divided according to the fixed time length; the mean and standard deviation of the magnetic field gradient change rate are calculated in each window by sliding window statistical analysis method; the amplitude threshold is established by analyzing the statistical characteristics of the window data in different geological activity periods, and the value range is usually 0.1-1.0nT.

[0078] A dynamic amplitude filtering algorithm is performed by the enhanced pulse timing sequence to eliminate redundant pulses below an amplitude threshold, generating a preliminary screened pulse sequence;

[0079] Further, the enhanced pulse timing sequence generated by Daubechies4 wavelet decomposition is first read, and the amplitude value and firing timestamp of each pulse event are analyzed; the pulse amplitude value is compared with a dynamic amplitude threshold defined based on historical geomagnetic gradient monitoring data in real time to establish a pulse validity determination condition; a sliding window detection mechanism is used to scan all pulses within a 10 ms time window, and redundant pulses with an amplitude value below the dynamic amplitude threshold are eliminated; the remaining pulse events are reordered according to the original timestamp to generate a preliminary screened pulse sequence.

[0080] The preliminary screened pulse sequence is phase calibrated by a voltage-controlled oscillator of a digital phase-locked loop, and a timing-aligned pulse sequence is output;

[0081] Further, the digital phase-locked loop first detects the phase deviation of the input pulse sequence, and the voltage-controlled oscillator dynamically adjusts the oscillation frequency according to the error signal output by the phase detector; when a pulse timing deviation is detected, the voltage-controlled oscillator precisely adjusts the phase of the output pulse by changing the control voltage, and the adjustment step is 0.1 ns each time; after 5-10 cycles of continuous phase tracking and correction, the timing jitter of the pulse sequence is controlled within ±2 ns; the finally output timing-aligned pulse sequence maintains the amplitude characteristics and waveform features of the original pulse, only the firing time is fine-tuned, and the rising edges of all pulses are strictly aligned with the reference clock edge. During the phase calibration process, the bandwidth parameter of the digital phase-locked loop is adaptively adjusted according to the input pulse rate, and the bandwidth range is set to 100 kHz-1 MHz to balance the tracking speed and stability.

[0082] A pulse interval compression algorithm is performed on the timing-aligned pulse sequence to improve the information density by adjusting the firing frequency, generating a high-density pulse sequence;

[0083] Further, the original firing interval distribution characteristics of the timing-aligned pulse sequence are first analyzed, and a pulse interval histogram is established; the maximum allowed compression ratio is determined based on the histogram statistics, and a nonlinear time scaling algorithm is used to gradually reduce the pulse interval under the premise of maintaining the integrity of the pulse waveform; the pulse amplitude distortion degree is dynamically monitored during the compression process, and the current compression operation is automatically stopped when the distortion degree exceeds 5%; the compressed pulse sequence is phase recalibrated by a digital phase-locked loop to eliminate the timing jitter introduced by time scaling; and the finally generated high-density pulse sequence maintains the amplitude characteristics and waveform integrity of the original pulse. The compression algorithm is implemented using a sliding window, and the window size is adaptively adjusted according to the pulse rate, with a range of 10-50 pulse periods.

[0084] The high-density pulse sequence is subjected to timing consistency optimization by a dynamic time warping algorithm to generate an optimized pulse sequence with high signal-to-noise ratio.

[0085] Further, firstly, a time domain correspondence relationship between the high-density pulse sequence and a reference template sequence is established, and the dynamic time warping algorithm calculates a minimum cumulative path distance of each pulse event and the template sequence; the algorithm adjusts the pulse firing time through nonlinear time warping to eliminate time distortion introduced by pulse interval compression; in the optimization process, a dynamic programming method is used to solve an optimal time warping path, and the constraint conditions include that a maximum time offset is not more than 50 ns and a path slope is kept within a range of 0.5-2.0; the pulse sequence subjected to timing adjustment is subjected to final phase calibration by a digital phase-locked loop to ensure that rising edges of all pulses are strictly synchronized with a system clock; and finally, an optimized pulse sequence generated keeps amplitude characteristics and waveform integrity of the original pulse. A window length of the dynamic time warping algorithm is adaptively adjusted according to a pulse density, and the range is set to 10-30 pulse periods.

[0086] The template sequence refers to a typical pulse mode reference sequence generated by historical data clustering, contains standardized time stamps and amplitude characteristics, and is used for timing alignment calculation in the dynamic time warping.

[0087] The minimum cumulative path distance of each pulse event and the template sequence is calculated, and the expression is as follows:

[0088]

[0089] wherein d n represents a minimum cumulative alignment distance of the first n events of the input pulse sequence and the template sequence, α is a time dimension weighting coefficient (usually, the value range is: 10 4 ~ 10 8 ), β is an amplitude dimension weighting coefficient (usually, the value range is: 0.1~1.0), t m is a time stamp of the mth event of the input sequence, M is a total event number of the template sequence, l is a current event index of the template sequence, is a template event time stamp aligned with the mth input event on the path π, a m is a normalized amplitude of the mth event of the input pulse sequence, is a template event amplitude aligned with the mth input event on the path π, is a set of all valid alignment paths from the starting point (1, 1) to the current point (n, l).

[0090] It should be noted that t m and are subjected to normalization processing before participating in calculation, so as to eliminate time-amplitude dimension differences and ensure distance calculation numerical stability.

[0091] S4, encapsulate the optimized pulse sequence into a standard transmission frame and transmit it to the seismic detection center, and simultaneously decode and reconstruct it through the CNN-LSTM model to generate high-fidelity seismic geomagnetic observation data.

[0092] A protocol header is added to the optimized pulse sequence, and the ASN.1 encoding format is used for encapsulation to generate a standard transmission frame.

[0093] Further, a 16-byte protocol header is added in front of the optimized pulse sequence, which includes a 4-byte synchronization word, a 4-byte sequence length identifier, a 4-byte timestamp, and a 4-byte checksum. The protocol header and the optimized pulse sequence are structured and encapsulated using the ASN.1 Basic Encoding Rules (BER), and are encoded field by field in the T-L-V (Type-Length-Value) triple format. During the encapsulation process, each sample point of the pulse sequence is encoded using the ASN.1 REAL type, and the timestamp is encoded using the ASN.1 GeneralizedTime type. The data unit after encoding is generated by the CRC-32 algorithm to generate a 4-byte frame check sequence, which is appended to the ASN.1 encoded data to form a complete standard transmission frame. The total length of the standard transmission frame dynamically changes according to the number of sample points of the optimized pulse sequence, with a minimum frame length of 64 bytes and a maximum of no more than 1024 bytes.

[0094] The standard transmission frame is transmitted to the seismic detection center through LPWAN, and the local spatial features in the standard transmission frame are extracted through the CNN convolutional layer of the CNN-LSTM model, and the long-term temporal dependence relationship between the local spatial features is learned through the gating mechanism of the LSTM network.

[0095] Further, the CNN convolutional layer of the CNN-LSTM model uses a three-layer convolution kernel structure to process the pulse sequence. The first layer uses 16 5×1 convolution kernels to extract local spatial features, the second layer uses 32 3×1 convolution kernels to capture higher-level spatial patterns, and the third layer uses 64 3×1 convolution kernels to refine the features. After each layer of convolution, a ReLU activation function and a max-pooling operation are connected, and finally a 128-dimensional spatial feature vector is output. The gating mechanism of the LSTM network processes the feature sequence output by the CNN, the input gate controls the feature update, the forget gate manages the long-term state memory, and the output gate adjusts the feature expression strength. The LSTM unit establishes long-term dependencies between local spatial features through recurrent connections.

[0096] The local spatial features include pulse waveform time-domain gradient features, three-axis sensor spatial coupling features, and multi-scale frequency domain energy features.

[0097] It should be noted that the CNN-LSTM model has a synergistic advantage compared to the traditional single-mode neural network, specifically: the CNN convolution layer extracts the time-domain gradient features and three-axis spatial coupling features of the pulse sequence through 5x1 / 3x1 multi-scale convolution kernels, and the LSTM network gating mechanism establishes a 50-sample time span to learn long-term dependencies, both of which synergistically optimize the spatio-temporal features, and finally output feature representations containing both waveform details and evolution laws.

[0098] Based on the long-term temporal dependence between local spatial features, the standard transmission frame is decoded and reconstructed, and the pulse timing inversion algorithm is used to recover the original waveform data with event markers.

[0099] Further, the spatio-temporal feature vector is decomposed into local spatial feature components and long-term temporal dependence feature components by principal component analysis; the local spatial feature components reconstruct the spatial distribution patterns of the three-axis seismic sensor and the three-axis geomagnetic sensor through inverse convolution operation, and the long-term temporal dependence feature components reconstruct the UTC timestamp marked pulse firing timing through the reverse propagation of the LSTM network; the pulse timing inversion algorithm fuses the reconstructed spatial distribution patterns and timing information, and through iterative optimization, the reconstructed pulse sequence matches the time-frequency features of the original pulse sequence, and finally restores the original waveform data with event markers.

[0100] The embodiment also provides a seismic geomagnetic observation data acquisition and transmission system, comprising: a data marking module for dynamically triggering a high-voltage pulse signal and generating an event marker data packet according to the environmental magnetic field gradient change rate;

[0101] A pulse synergistic enhancement module is used to acquire original waveform data with event markers, calculate synergistic pulse firing based on a pulse neural network, and obtain an enhanced pulse timing sequence;

[0102] A pulse optimization module is used to adopt a pulse filtering algorithm based on a dynamic amplitude threshold, combine a phase-locked loop synchronization method to optimize the information density and timing consistency of the enhanced pulse timing sequence, and generate an optimized pulse sequence with high signal-to-noise ratio;

[0103] A seismic waveform reconstruction module is used to encapsulate the optimized pulse sequence into a standard transmission frame and transmit it to a seismic detection center, and simultaneously decode and reconstruct it through a CNN-LSTM model to generate high-fidelity seismic geomagnetic observation data.

[0104] The embodiment also provides a computer device suitable for the seismic geomagnetic observation data acquisition and transmission method, comprising: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the seismic geomagnetic observation data acquisition and transmission method proposed in the above embodiment.

[0105] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved by WIFI, an operator network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0106] The embodiment also provides a storage medium having a computer program stored thereon, the program being executed by a processor to implement the method for collecting and transmitting seismic geomagnetic observation data according to the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage, a flash memory, a magnetic disk or an optical disk.

[0107] To sum up, the present application achieves the following effects: through the pulse neural network computing environment, the biological characteristics of the LIF neuron are used to realize the pulse coordinated firing of the original waveform data with event markers, and the information expression capability of the pulse time sequence is effectively enhanced; through the CNN-LSTM model, the spatial features of the transmission frame are extracted by the CNN convolution layer, the time sequence dependency is learned by combining the LSTM network, and the precise reconstruction of the compressed transmission data is realized. The two steps work together to form a complete technical chain from data collection to transmission reconstruction, which not only guarantees the real-time response capability of the geomagnetic anomaly event, but also ensures the integrity and accuracy of the observation data.

[0108] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A method for seismic geomagnetic observation data acquisition transmission, characterized in that: The application relates to a high-sensitivity and high-fidelity seismic geomagnetic observation method and device. According to the environmental magnetic field gradient change rate, a high-voltage pulse signal is dynamically triggered, and an event marking data packet is generated; Raw waveform data with event marking is collected, a cooperative pulse firing is performed based on a pulse neural network computing environment, and an enhanced pulse timing sequence is obtained; An enhanced pulse timing sequence information density and timing consistency are optimized by adopting a pulse filtering algorithm based on a dynamic amplitude threshold and a phase-locked loop synchronization method, and an optimized pulse sequence with high signal-to-noise ratio is generated; The optimized pulse sequence is packaged into a standard transmission frame and transmitted to a seismic detection center, and meanwhile, the CNN-LSTM model is decoded and reconstructed to generate high-fidelity seismic geomagnetic observation data.

2. The method for seismic geomagnetic observation data collection transmission according to claim 1, characterized in that: The step of dynamically triggering a high-voltage pulse signal according to the environmental magnetic field gradient change rate and generating an event marking data packet is as follows, The environmental magnetic field gradient change rate is continuously monitored through a Terfenol-D alloy sheet; Based on historical geomagnetic gradient data statistical analysis, an effective event trigger threshold is defined; When the detected environmental magnetic field gradient change rate is greater than or equal to the effective event trigger threshold, the magnetostrictive effect trigger signal is used to drive the Terfenol-D alloy sheet to produce deformation, and the PZT-5H piezoelectric ceramic is excited through mechanical coupling to output a high-voltage pulse signal; After the high-voltage pulse signal is excited, the UTC time stamp, the instantaneous magnetic field intensity change rate and the background magnetic field baseline are synchronously acquired, and the TLV format is packaged into an event marking data packet through an SPI interface.

3. The method for seismic geomagnetic observation data collection transmission according to claim 1, characterized in that: The raw waveform data with event marking is collected by synchronously collecting raw waveform data through a three-axis seismic sensor and a three-axis geomagnetic sensor, and the event marking data packet is associated in real time during the collection process to generate raw waveform data with event marking. The raw waveform data with event marking includes three-axis seismic waveform, three-axis geomagnetic waveform, UTC time stamp, instantaneous magnetic field intensity change rate and background magnetic field baseline.

4. The method for seismic geomagnetic observation data collection transmission of claim 3, wherein: The step of performing cooperative pulse firing based on a pulse neural network computing environment to obtain an enhanced pulse timing sequence is as follows, An Intel Loihi 2 neuromorphic chip is adopted to initialize a pulse neural network computing environment composed of LIF neurons; Based on the pulse neural network computing environment, the LIF neurons are used to analyze the initial waveform data with time marking, and when the dynamic event trigger threshold is exceeded, cooperative pulse firing is triggered to generate original pulse signals; A three-level decomposition algorithm based on Daubechies4 wavelet is used to perform multi-scale analysis on the firing timing and spatial distribution of the original pulse signals, uniform quantization coding is performed after the time-frequency feature coefficients are extracted, and an enhanced pulse timing sequence is generated.

5. The method for seismic geomagnetic observation data collection transmission of claim 1, wherein: The step of optimizing the information density and timing consistency of the enhanced pulse timing sequence by adopting a pulse filtering algorithm based on a dynamic amplitude threshold and a phase-locked loop synchronization method to generate an optimized pulse sequence with high signal-to-noise ratio is as follows, Based on historical geomagnetic gradient monitoring data, an amplitude threshold is defined through a sliding window statistical analysis method; A dynamic amplitude filtering algorithm is performed on the enhanced pulse timing sequence to remove redundant pulses below the amplitude threshold, and a preliminarily screened pulse sequence is generated. The pulse sequence after the preliminary screening is phase calibrated by a voltage-controlled oscillator of a digital phase-locked loop, and a time-aligned pulse sequence is output; An inter-pulse compression algorithm is performed on the time-aligned pulse sequence, the information density is improved by adjusting the firing frequency, and a high-density pulse sequence is generated; A dynamic time warping algorithm is used to perform timing consistency optimization on the high-density pulse sequence, and an optimized pulse sequence with high signal-to-noise ratio is generated.

6. The method for seismic geomagnetic observation data collection transmission of claim 1, wherein: The optimized pulse sequence is encapsulated into a standard transmission frame, which means adding a protocol header to the optimized pulse sequence and encapsulating it in ASN.1 encoding format to generate a standard transmission frame.

7. The method for seismic geomagnetic observation data collection transmission of claim 6, wherein: The decoding and reconstruction by the CNN-LSTM model generates high-fidelity seismic geomagnetic observation data, and the steps are as follows, The standard transmission frame is transmitted to the seismic detection center through LPWAN, the local spatial features in the standard transmission frame are extracted through the CNN convolution layer of the CNN-LSTM model, and the long-term temporal dependence relationship between local spatial features is learned through the gating mechanism of the LSTM network; Based on the long-term temporal dependence relationship between local spatial features, the standard transmission frame is decoded and reconstructed, and the original waveform data with event markers are recovered by using the pulse timing inversion algorithm.

8. A system for seismic geomagnetic observation data acquisition and transmission, based on the method for seismic geomagnetic observation data acquisition and transmission according to any one of claims 1 to 7, characterized in that: It comprises, The data marking module is used to dynamically trigger high-voltage pulse signals and generate event marker data packets according to the environmental magnetic field gradient change rate; The pulse synergistic enhancement module is used to collect original waveform data with event markers, calculate the synergistic pulse firing of the environment based on the pulse neural network, and obtain enhanced pulse timing sequences; The pulse optimization module is used to adopt a pulse filtering algorithm based on a dynamic amplitude threshold, and combine a phase-locked loop synchronization method to optimize the information density and timing consistency of the enhanced pulse timing sequence, and generate an optimized pulse sequence with high signal-to-noise ratio; The seismic waveform reconstruction module is used to encapsulate the optimized pulse sequence into a standard transmission frame and transmit it to the seismic detection center, and simultaneously decode and reconstruct it by the CNN-LSTM model to generate high-fidelity seismic geomagnetic observation data. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to realize the steps of the method for collecting and transmitting seismic geomagnetic observation data according to any one of claims 1-7.

10. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the method for collecting and transmitting seismic geomagnetic observation data according to any one of claims 1-7.

Images