Landslide surge wave height inversion method, system, device and medium based on distributed optical fiber sensing

Abstract

This invention discloses a method, apparatus, and electronic device for inverting landslide surge wave height based on distributed optical fiber sensing. The method includes: acquiring the raw time-series signal collected by an optical fiber sensing array and performing adaptive denoising processing; separating high-frequency landslide vibration and low-frequency surge components using time-frequency analysis technology to construct a spatiotemporal evolution spectrum; accurately determining landslide surge events based on the distinct trajectory slope characteristics exhibited in the spectrum due to the significant difference in propagation speed between the two components; constructing a frequency-domain phase-pressure conversion model, introducing a frequency-dependent correction factor to restore the true surge pressure spectrum, and determining landslide deformation correction coefficients based on wave train superposition characteristics to compensate for energy dissipation; and finally, combining the above factors and coefficients to invert and obtain the corrected landslide surge wave height. This invention effectively solves the problems of difficult signal identification and insufficient inversion accuracy in complex water areas through physical field decoupling and a dual correction mechanism, achieving high-precision quantitative monitoring of the entire landslide surge process.

Description

Technical Field

[0001] This invention relates to the field of geological disaster monitoring and fiber optic sensing technology, specifically to a landslide surge monitoring method, system, electronic device, and computer-readable storage medium based on distributed fiber optic sensing. Background Technology

[0002] Landslide surges are a highly destructive secondary disaster that occurs when a landslide on a reservoir bank becomes unstable and enters the water. The massive surges generated by the rapid entry of the landslide into the water not only erode the opposite bank and downstream river channels, but in severe cases, they can also create a "shock wave" effect, posing a significant threat to dam safety and the lives and property of residents along the riverbank. Therefore, accurately obtaining dynamic parameters such as wave height and propagation speed of the surges is a core basis for disaster early warning and risk assessment.

[0003] Traditional landslide surge monitoring mainly relies on point sensors such as wave height meters and pore water pressure gauges. However, these contact-based point monitoring methods have obvious limitations: first, the spatial coverage is limited, making it difficult to capture the continuous evolution of surges over long river channels; second, the equipment is easily damaged by siltation and impacts from floating debris in harsh outdoor aquatic environments; and third, the real-time and synchronous nature of the data is poor, making it difficult to meet the needs of rapid early warning for sudden disasters.

[0004] In recent years, distributed fiber acoustic sensing (DAS) technology, especially ultra-weak fiber Bragg grating (uwDAS) technology, has been increasingly applied to underwater geological monitoring due to its long-distance, interference-resistant, and fully distributed characteristics. Existing DAS-based monitoring schemes typically utilize fiber optic cables laid on the seabed to sense fiber optic strain caused by changes in water pressure, attempting to infer surge parameters from this.

[0005] However, existing DAS inversion techniques still face severe technical bottlenecks in the specific scenario of landslide surges, mainly in the following three aspects: Signal aliasing and identification difficulties: The landslide surge process is a complex solid-liquid-fiber coupling process. The original fiber optic signal not only contains low-frequency pressure fluctuations caused by the surge, but also high-frequency vibrations from the landslide impacting the riverbed, water flow turbulence noise, and environmental thermal noise. Current technologies lack effective time-frequency separation methods, making it difficult to accurately extract effective signals characterizing the surge dynamics from the high-noise background, resulting in a high false alarm rate and the inability to utilize the propagation characteristics of different physical fields (vibration and water waves) for event cross-verification.

[0006] Linearization error in the conversion model: Most existing inversion methods use fixed, linear strain-pressure conversion coefficients (i.e., assuming that fiber strain is proportional to water pressure). However, actual underwater acoustic physics shows that the response sensitivity of optical fibers to water waves of different frequencies varies significantly (i.e., frequency dependence exists). Ignoring the frequency response function and directly using constants for calculation will lead to waveform distortion and large errors in wave height calculation.

[0007] Ignoring the energy dissipation of the source mechanism: Existing models typically assume that the landslide body is a rigid body, believing that the landslide potential energy is completely converted into surge wave energy. However, in actual geological disasters, the landslide body undergoes violent fragmentation, deformation, and disintegration during its entry into water, a process (i.e., landslide deformation) that consumes a significant amount of energy. If existing technologies do not consider this energy dissipation of the source mechanism, they often significantly overestimate the initial wave height of the surge, leading to excessively high warning thresholds, causing unnecessary panic or engineering waste. Summary of the Invention

[0008] To address the aforementioned deficiencies in existing technologies, this invention provides a method, system, device, and medium for landslide surge wave height inversion based on distributed optical fiber sensing. The aim is to solve the technical problems in existing DAS monitoring technologies, such as difficulty in signal identification, low inversion accuracy due to nonlinear frequency response, and large wave height prediction deviations caused by neglecting landslide deformation.

[0009] To address the above problems, a first aspect of the present invention provides a method for inverting landslide surge wave height based on distributed optical fiber sensing, comprising: The raw time-series signals collected by the distributed fiber optic sensor array laid on the bottom of the water body are acquired and then denoised and filtered. The processed signal was decomposed using time-frequency analysis technology to separate the high-frequency landslide vibration component and the low-frequency surge component, and the spatiotemporal evolution spectrum of the two components was constructed. Based on the velocity difference characteristics of landslide vibration and surge propagation in the spatiotemporal evolution map, landslide surge events are identified and surge propagation velocity is calculated; By combining frequency-dependent correction and landslide deformation correction, the low-frequency surge component is inverted to obtain wave height parameters, including: A frequency-dependent correction factor is determined based on the frequency response characteristics of fluid-optical fiber interaction, and the phase signal of the low-frequency surge component is mapped to hydrodynamic pressure data using the frequency-dependent correction factor. Using the landslide deformation correction coefficient determined based on wave train superposition characteristics, the wave height retrieved from the hydrodynamic pressure data is corrected for energy dissipation, thus obtaining the final landslide surge wave height.

[0010] Furthermore, in the above method, the denoising and filtering process on the original time-series signal includes: The original time-series signal is processed using an adaptive wavelet threshold denoising algorithm, and the threshold is dynamically adjusted according to the signal-to-noise ratio: a soft threshold algorithm is used when the signal-to-noise ratio is less than 10dB, and a hard threshold algorithm is used when the signal-to-noise ratio is greater than or equal to 10dB. The denoised signal is filtered using a directional bandpass filter, wherein the passband frequency range of the directional bandpass filter is 0.04Hz to 50Hz.

[0011] Furthermore, in the above method, the decomposition of the processed signal using time-frequency analysis technology includes: performing convolution operation on the processed signal using Morlet wavelet transform to convert the one-dimensional time-series signal into a two-dimensional time-frequency domain signal; based on the frequency distribution characteristics of the two-dimensional time-frequency domain signal, setting a frequency segmentation threshold, extracting components with frequencies higher than the threshold as the high-frequency landslide vibration component, and extracting components with frequencies lower than the threshold and located in the 0.04Hz to 15Hz frequency band as the low-frequency surge component.

[0012] Furthermore, in the above method, the step of determining the landslide surge event and calculating the surge propagation velocity based on the velocity difference characteristics of landslide vibration and surge propagation in the spatiotemporal evolution map includes: Identify the near-vertical trajectory formed by the high-frequency landslide vibration component in the spatiotemporal evolution map, and take its initial arrival time as the starting zero point of the landslide water entry event; Identify the inclined trajectory formed by the low-frequency surge component in the spatiotemporal evolution map, and calculate the propagation speed of the surge by least squares fitting based on the slope of the inclined trajectory in the spatiotemporal domain or the time difference with the starting zero point. The criteria for determining a landslide surge event are: simultaneously detecting the near-vertical trajectory and the inclined trajectory, whose propagation speeds differ by more than a preset threshold, within the same spatiotemporal region.

[0013] Furthermore, in the above method, the step of determining a landslide surge event further includes: Extract the spatial energy distribution characteristics of the high-frequency landslide vibration component along the fiber optic sensing array; If the spatial energy distribution characteristics satisfy the bidirectional exponential decay model centered on the landslide entry point, then the monitored event is confirmed as a valid landslide surge event. The bidirectional exponential decay model characterizes the physical property that vibration energy decays faster in the near field than in the far field.

[0014] Furthermore, in the above method, determining the frequency-dependent correction factor based on the frequency response characteristics of the fluid-optical fiber interaction includes: Power spectrum analysis is performed on the low-frequency surge component to obtain the power spectral density at different frequencies; based on the ratio of the power spectral density to the preset standard fluid pressure spectral density, the frequency dependence correction factor that varies with frequency is calculated. The fiber frequency domain strain signal and its dependent correction factor were calculated. The relationship between them, the calculation formula satisfies:

[0015] in, It is a fiber optic frequency domain strain signal. This is the proportionality coefficient. The dynamic pressure generated by the surge.

[0016] Furthermore, in the above method, the step of using the landslide deformation correction coefficient determined based on wave train superposition characteristics to correct the wave height derived from the hydrodynamic pressure data for energy dissipation includes: The number of wave crests and the number of wave train superpositions whose peak amplitudes exceed a preset threshold are counted in the time sequence of the low-frequency surge component. Based on the number of wave train superpositions, a landslide deformation correction coefficient is determined to characterize the degree of landslide fragmentation upon entering the water and the proportion of energy dissipation. ; Introducing landslide deformation correction factor Corrected landslide surge wave height The calculation formula satisfies:

[0017] in, For water density, Because of the water depth, The angle between the surge crest line and the fiber axis.

[0018] According to another aspect of the present invention, the present invention also provides a landslide surge wave height inversion system based on distributed optical fiber sensing, comprising: The signal acquisition module is used to acquire the raw time-series signals collected by the distributed optical fiber sensor array laid on the bottom of the water body; The preprocessing and time-frequency decomposition module is used to denoise and filter the original time-series signal, and to separate the high-frequency landslide vibration component and the low-frequency surge component using time-frequency analysis technology. The event determination and velocity calculation module is used to construct a spatiotemporal evolution map, determine landslide surge events and calculate surge propagation velocity based on the velocity difference characteristics between landslide vibration and surge propagation; The parameter inversion module is used to combine frequency-dependent correction and landslide deformation correction to map the low-frequency surge component into hydrodynamic pressure and correct energy dissipation, and output landslide surge wave height parameters.

[0019] According to another aspect of the present invention, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method as described in any one of the first aspects.

[0020] According to another aspect of the present invention, a computer-readable storage medium is also provided thereon storing a computer program, characterized in that the computer program, when executed by a processor, implements the method as described in any one of the first aspects.

[0021] Existing technologies often struggle to distinguish between landslide surges and shipping waves or underwater earthquakes. This invention decouples the original signal into high-frequency landslide vibration components and low-frequency surge components through time-frequency analysis. Utilizing the physical fact that their propagation speeds in water differ by approximately two orders of magnitude, it captures the distinct trajectory slope characteristics (i.e., coexistence of vertical and inclined trajectories) exhibited in their spatiotemporal evolution. This determination mechanism leverages the unique source mechanism of landslides entering water, which simultaneously excites acoustic and gravity waves. This fundamentally differentiates it from ships that only excite gravity waves (resulting in only inclined trajectories) or earthquakes that only excite elastic waves (resulting in only vertical trajectories with no center), thereby significantly reducing the false alarm rate. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the landslide surge monitoring device in this invention; Figure 2 This is a flowchart of Embodiment 1 of the present invention; Figure 3 A schematic diagram illustrating the spatiotemporal evolution of landslide surge propagation is shown. Figure 4 The diagram schematically illustrates the propagation and attenuation process of the inverted landslide vibration. Figure 5 A schematic diagram illustrating the wave propagation process based on water level; Figure 6 This is a structural block diagram of Embodiment 2 of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0024] The specific implementation of the embodiments of this application will be further described below with reference to the accompanying drawings.

[0025] Figure 1An exemplary landslide surge monitoring device applicable to embodiments of this application is shown. For example... Figure 1 As shown, the device 10 may include a distributed optical fiber sensor array 12 laid on the bottom of the water, an optical fiber signal demodulation device 14, and a data processing terminal 16.

[0026] It should be noted that the solutions in this application embodiment can be executed either in an integrated fiber optic demodulator or by the demodulator acquiring data and transmitting it to a backend data processing terminal (such as a server, industrial control computer, or cloud) for collaborative execution. The distributed fiber optic sensor array 12 can be a sensor network built based on ultra-weak fiber Bragg grating (uwDAS) technology, laid at the bottom of rivers, reservoirs, or lakes to sense vibrations and pressure changes in the aquatic environment. The fiber optic signal demodulation device 14 is used to emit pulsed light into the fiber and receive Rayleigh scattering or Bragg reflection signals, demodulating them into time-series phase or strain signals.

[0027] The data processing terminal 16 can be any suitable device with computing capabilities, including but not limited to high-performance servers, workstations, or embedded computing modules. The data processing terminal 16 runs the landslide surge monitoring algorithm program described in this application embodiment, used to denoise, decompose, extract features, and perform inversion calculations on the collected raw signals to obtain accurate dynamic parameters for the landslide surge event in the water area.

[0028] Based on the above system, this application provides a landslide surge wave height inversion method based on distributed optical fiber sensing, which will be described in detail below through several embodiments.

[0029] Firstly, this application provides a method for inverting landslide surge wave height based on distributed optical fiber sensing.

[0030] refer to Figure 2 This paper illustrates a landslide surge wave height inversion method based on distributed optical fiber sensing according to Embodiment 1 of this application. The monitoring method includes the following steps: Step S100: Acquire the raw time-series signal collected by the distributed optical fiber sensor array laid on the bottom of the water area, and perform noise reduction and filtering processing on the raw time-series signal.

[0031] Due to the highly complex underwater acoustic environment, the raw signals acquired by optical fibers are often mixed with background thermal noise, water flow turbulence noise, and the optical path noise of the system itself. Directly using the raw signals for inversion will result in an extremely low signal-to-noise ratio. Therefore, high-quality data is required.

[0032] In a preferred embodiment of this application, in order to balance the preservation of weak signals and the removal of strong noise, the denoising and filtering process specifically adopts the following steps: First, an adaptive wavelet threshold denoising algorithm is used to process the original time-series signal. Considering the non-stationary and transient characteristics of landslide surge signals, this embodiment introduces the signal-to-noise ratio (SNR) as a decision criterion for dynamic adjustment. When the signal-to-noise ratio of the monitored signal is less than 10 dB (typically corresponding to weak signals during the propagation of surge waves to the far field or the initial stage of landslide entry into water), a soft thresholding algorithm is used. Soft thresholding has good continuity and can avoid artifacts near the signal zero point, thereby preserving the characteristics of weak surge signals to the greatest extent.

[0033] When the signal-to-noise ratio of the monitored signal is greater than or equal to 10dB (typically corresponding to the arrival of the main surge or strong vibration signal), a hard thresholding algorithm is used. Hard thresholding can thoroughly filter out background noise and maintain the steepness of the signal edges, which is beneficial to improving the accuracy of subsequent first arrival time acquisition.

[0034] Secondly, a directional bandpass filter is used to constrain the frequency domain of the denoised signal. In this embodiment, the preferred passband frequency range is 0.04Hz to 50Hz. Setting 0.04Hz as the lower limit is to filter out low-frequency baseline drift caused by slow changes in ambient temperature and to prevent the inversion baseline from tilting; setting 50Hz as the upper limit is to filter out high-frequency flow noise caused by water turbulence and to ensure that the signal energy is concentrated in the effective frequency band of landslide vibration and surge fluctuations.

[0035] Step S200: Decompose the processed signal using time-frequency analysis technology to separate the high-frequency landslide vibration component and the low-frequency surge component, and construct the spatiotemporal evolution spectrum of the two components.

[0036] The landslide surge process involves two distinct physical signals: the first is the solid-liquid coupling vibration (high frequency, high wave velocity) generated by the landslide impacting the riverbed, and the second is the gravity wave surge (low frequency, low wave velocity) generated by the undulation of the water surface. These two signals are highly superimposed on the original time-domain waveform and must be separated using time-frequency analysis techniques in order to utilize their respective characteristics.

[0037] In a preferred embodiment of this application, Morlet wavelet transform is used as the time-frequency analysis tool. The waveform of the Morlet wavelet is highly similar to the underwater acoustic impact response, and it has both good time-domain and frequency-domain resolution. Through convolution operations, the one-dimensional time-series signal is mapped into a two-dimensional time-frequency domain signal.

[0038] Based on the physical characteristics of landslide surge waves, this embodiment sets 15Hz as the frequency segmentation threshold. Specifically, it includes: extracting components with frequencies below 15Hz and above 0.04Hz as low-frequency surge wave components, which characterize the long-period fluctuation of water pressure and serve as the basis for subsequent wave height inversion; and extracting components with frequencies above 15Hz as high-frequency landslide vibration components, which characterize the elastic waves generated by the impact of rock and soil and serve as the time reference for subsequent determination of the event occurrence time.

[0039] Subsequently, the two separated components are spread out along the fiber optic cable laying distance to construct a spatiotemporal evolution map (i.e., a waterfall diagram, if...). Figure 3 As shown in the figure, to visually represent the signal over time. and space The evolutionary trajectory.

[0040] Step S300: Based on the velocity difference characteristics of landslide vibration and surge propagation and the energy distribution law (e.g., energy attenuation characteristics) in the spatiotemporal evolution map, determine the landslide surge event and calculate the surge propagation velocity.

[0041] By utilizing the differences in the propagation speed of different physical waves, accurate event identification and self-calibration can be achieved, solving the problem of high false alarm rate in single face recognition.

[0042] As a preferred embodiment of this application, it specifically includes: In the spatiotemporal evolution map, two types of characteristic trajectories are identified: near-vertical trajectories and inclined trajectories. The near-vertical trajectories are formed by high-frequency landslide vibration components. Because the propagation speed of sound waves in water or elastic waves in sediments is extremely fast (typically >1400 m / s), they appear as nearly vertical lines within a limited spatiotemporal window. In this embodiment, the initial arrival time is marked as the absolute starting point zero of the landslide entering the water event. The inclined trajectory is formed by low-frequency surge components. Because gravity waves propagate slowly (usually <30m / s), they appear as an oblique line extending downstream over time.

[0043] When both near-vertical and inclined trajectories are detected simultaneously within the same spatiotemporal region, and the difference in their propagation speeds exceeds a preset threshold (e.g., a difference of more than a hundred times), further energy verification is performed: The processor extracts the spatial energy distribution of the vibration component corresponding to the near-vertical trajectory along the optical fiber. If this energy distribution satisfies a bidirectional exponential decay model centered on the landslide entry point (i.e., fast near-field decay and slow far-field decay), then a landslide surge event is determined to have occurred. The dual verification logic of "velocity difference + energy decay" effectively eliminates simple shipping interference (only inclined lines), far-field seismic interference (only vertical lines with no obvious central decay characteristics), or electromagnetic interference (uniform energy distribution along the entire line).

[0044] Then, using the identified starting zero point ( Based on the baseline, the arrival time of each spatial point on the inclined trajectory is extracted, and the linear regression equation of the spatiotemporal data is fitted by the least squares method. The slope of the equation is the high-precision wave propagation speed.

[0045] Step S400: Combining frequency-dependent correction and landslide deformation correction, the low-frequency surge component is inverted to obtain wave height parameters.

[0046] By introducing a dual correction mechanism, the inversion error problem of traditional linear models under non-uniform media and complex source mechanisms is solved.

[0047] This step specifically includes the following steps: Step S401: Since the fiber optic sensing cable does not rigidly sense all frequencies of water waves, its response sensitivity to signals of different frequencies varies. As a preferred embodiment, this example introduces a frequency-dependent correction factor. .

[0048] Specifically, by performing power spectrum analysis on the low-frequency surge component, the ratio of its measured power spectral density to the standard fluid pressure spectral density is calculated, thus obtaining the dependent correction factor. Furthermore, the frequency domain strain signal of the optical fiber and its dependence on the correction factor were calculated. The relationship between them, the calculation formula satisfies:

[0049] in, It is a fiber optic frequency domain strain signal. The proportionality coefficient (taken as 2 × 10⁻⁶ in this experiment) -10 Pa -1 ), The dynamic pressure generated by the surge; Step S402: Considering that the landslide body is not a completely rigid body during the water entry process, its fragmentation, disintegration, and deformation processes will consume some potential energy, resulting in the generated surge wave height being lower than the theoretical value. If no correction is made, the inversion result will be too high. As a preferred implementation, this embodiment extracts wave train superposition features to quantify this effect.

[0050] Considering static pressure, total pressure It can be decomposed into static pressure and the dynamic pressure generated by the swell The formula is as follows:

[0051] in, Surge amplitude ( , (for high swell waves) For water density, Because of the water depth, The angle between the surge crest line and the fiber axis; Next, the number of peaks in the low-frequency surge component over time is counted. If the waveform exhibits a multi-peak superposition pattern (i.e., the number of wave train superpositions), then... An increase in the number of wave train superpositions indicates a high degree of landslide fragmentation. Based on this wave train superposition, a landslide deformation correction coefficient is determined. ( Typically based on the number of wave train superpositions calculate, ).

[0052] Introducing landslide deformation correction factor Corrected landslide surge wave height , The calculation formula is: .

[0053] Through the above steps, this embodiment not only achieves qualitative identification of landslide surge events, but also achieves high-precision quantitative inversion of dynamic parameters such as wave height and velocity.

[0054] By introducing an energy distribution verification mechanism based on a bidirectional exponential decay model, this application overcomes the limitations of existing monitoring technologies that rely solely on signal arrival time (velocity differences). These technologies are susceptible to electromagnetic interference from optical systems or large-scale geological vibrations (such as earthquakes) in the far field, which can manifest as high-speed vertical lines on spatiotemporal maps. This application utilizes the physical characteristics of landslide impact point sources to verify whether the energy exhibits a "high in the middle, exponential decay at both ends" pattern. This effectively eliminates electromagnetic interference with uniform energy distribution along the entire line or background noise lacking central decay characteristics, ensuring the authenticity of alarm signals and significantly improving anti-interference capabilities and false alarm rejection rates. Furthermore, by fitting a bidirectional exponential decay curve, the energy peak center (i.e., the landslide entry point into water) can be pinpointed. This not only complements the spatiotemporal trajectory differentiation characteristics based on propagation velocity differences but also verifies the physical event generation mechanism from an energy perspective. This ensures that the monitoring results fully conform to the physical laws of fluid-structure interaction dynamics, enhancing the physical reliability of locating the landslide entry point into water.

[0055] In the initial stage of the monitoring process, the processor first performs denoising and time-frequency decomposition on the acquired raw time-series signal, separating high-frequency and low-frequency components, and constructing an evolutionary map in the spatiotemporal domain, such as... Figure 3 As shown, the landslide event is determined by the difference in velocity between landslide vibration and surge propagation.

[0056] refer to Figure 4The vibration signal generated by the landslide entering the water is shown in the spatiotemporal diagram as the landslide vibration energy attenuates with the propagation distance. Linear fitting using the linear square method shows that the signal propagates extremely quickly (approximately 1333 m / s in the experimental water area). From... Figure 3 Data from region A is extracted for time-frequency analysis, such as... Figure 4 As shown, the vibration energy exhibits a significant exponential decay characteristic with increasing propagation distance. Combined with... Figure 3 Compared to Figure 4 High-speed vibration (in) Figure 3 The lower left corner is represented by a vertical line trajectory indicated by the red arrow. Figure 3 The low-frequency surge component shown in the image appears as a sloping colored area extending slowly downstream over time (i.e., the sloping trajectory indicated by the blue arrow). This sloping trajectory coexists with the aforementioned vertical trajectory in the same spatiotemporal domain. Due to the significant difference in their propagation speeds, they exhibit distinctly different trajectory slope characteristics, verifying the physical mechanism by which landslides simultaneously generate high-speed sound waves and low-speed water waves upon entering the water. By capturing this spatiotemporal trajectory differentiation characteristic, precise location of landslide events can be achieved.

[0057] After identifying the landslide event, the next step is to count the number of wave train superpositions to address the landslide fragmentation effect, referring to... Figure 4 and Figure 5 The degree of signal aliasing caused by boundary reflection is quantified by counting the number of superpositions of wave trains; otherwise, direct inversion will lead to errors in wave height calculation due to multi-wave interference. Figure 5 The green portion represents the calm period on the river surface before the swell arrives. The red portion reflects the propagation and superposition of the swell wave train. During propagation, the trough of the first wave becomes increasingly shallow; the wave becomes unstable and forms wave groups, with a longer period. The second and third waves superimpose, forming a complex wave train near wave trough meter #48. The red time domain and... Figure 4 The areas bounded by the blue and gray lines are consistent. The subsequent wave train (wake wave) exhibits significant oscillations and flattening starting from the 39# wave height meter position, and is superimposed with various reflected waves. Finally, a landslide deformation correction coefficient is introduced to correct for the inversion of the surge wave height.

[0058] It should be particularly emphasized that the specific parameters mentioned in the above embodiments (such as 10dB, 15Hz, 0.04-50Hz, etc.) are preferred values ​​based on specific experimental environments. In practical applications, technicians can make adaptive adjustments to the above parameters according to the hydrogeological conditions of specific water areas. As long as the core logic follows the technical concept of this application, they should all fall within the protection scope of this invention.

[0059] Secondly, this application also provides a landslide surge wave height inversion device based on distributed optical fiber sensing. Figure 6 shows a functional module architecture diagram of a landslide surge wave height inversion device based on distributed optical fiber sensing according to Embodiment 2 of this application. This architecture corresponds to the logical function division within the data processing terminal.

[0060] like Figure 3 As shown, the monitoring device 700 includes: Signal acquisition module 702: Used to communicate with the front-end fiber optic signal demodulation equipment to acquire the raw time-series signals collected by the distributed fiber optic sensor array laid on the bottom of the water body. This module may include a data buffer unit to handle data surges at high sampling rates.

[0061] Preprocessing and time-frequency decomposition module 704: This module is used to denoise and filter the original time-series signal, and to separate the high-frequency landslide vibration component and the low-frequency surge component using time-frequency analysis techniques. This module integrates an adaptive wavelet threshold algorithm library and a Morlet wavelet transform operation unit.

[0062] Event Determination and Velocity Calculation Module 706: This module is used to construct a spatiotemporal evolution map, determine landslide and surge events based on the velocity difference characteristics between landslide vibration and surge propagation, and calculate the surge propagation velocity. This module can identify near-vertical and inclined trajectories in a "K"-shaped feature and perform least-squares fitting calculations.

[0063] Parameter inversion module 708: This module combines frequency-dependent correction and landslide deformation correction to map the low-frequency surge component to hydrodynamic pressure and correct for energy dissipation, outputting landslide surge wave height parameters. This module stores a pre-calibrated table of frequency-dependent correction factors and a table of landslide deformation correction coefficients for executing the final physical inversion formula.

[0064] Thirdly, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in Embodiment 1.

[0065] It should be understood that the specific embodiments described above are merely illustrative or explanatory of the principles of the invention and do not constitute a limitation thereof. Therefore, any modifications, equivalent substitutions, improvements, etc., made without departing from the spirit and scope of the invention should be included within the protection scope of the invention. Furthermore, the appended claims are intended to cover all variations and modifications falling within the scope and boundaries of the appended claims, or equivalent forms of such scope and boundaries.

Claims

1. A method for inverting landslide surge wave height based on distributed optical fiber sensing, characterized in that, include: The raw time-series signals collected by the distributed fiber optic sensor array laid on the bottom of the water body are acquired and then denoised and filtered. The processed signal was decomposed using time-frequency analysis technology to separate the high-frequency landslide vibration component and the low-frequency surge component, and the spatiotemporal evolution spectrum of the two components was constructed. Based on the velocity difference characteristics of landslide vibration and surge propagation in the spatiotemporal evolution map, landslide surge events are identified and surge propagation velocity is calculated; By combining frequency-dependent correction and landslide deformation correction, the low-frequency surge component is inverted to obtain wave height parameters, including: A frequency-dependent correction factor is determined based on the frequency response characteristics of fluid-optical fiber interaction, and the phase signal of the low-frequency surge component is mapped to hydrodynamic pressure data using the frequency-dependent correction factor. Using the landslide deformation correction coefficient determined based on wave train superposition characteristics, the wave height retrieved from the hydrodynamic pressure data is corrected for energy dissipation, thus obtaining the final landslide surge wave height.

2. The method according to claim 1, characterized in that, The denoising and filtering process for the original time-series signal includes: The original time-series signal is processed using an adaptive wavelet threshold denoising algorithm, and the threshold is dynamically adjusted according to the signal-to-noise ratio: a soft threshold algorithm is used when the signal-to-noise ratio is less than 10dB, and a hard threshold algorithm is used when the signal-to-noise ratio is greater than or equal to 10dB. The denoised signal is filtered using a directional bandpass filter, wherein the passband frequency range of the directional bandpass filter is 0.04Hz to 50Hz.

3. The method according to claim 1, characterized in that, The decomposition of the processed signal using time-frequency analysis technology includes: performing convolution operations on the processed signal using Morlet wavelet transform to convert the one-dimensional time-series signal into a two-dimensional time-frequency domain signal; based on the frequency distribution characteristics of the two-dimensional time-frequency domain signal, setting a frequency segmentation threshold, extracting components with frequencies higher than the threshold as the high-frequency landslide vibration component, and extracting components with frequencies lower than the threshold and located in the 0.04Hz to 15Hz frequency band as the low-frequency surge component.

4. The method according to claim 1, characterized in that, The determination of landslide surge events and calculation of surge propagation velocity based on the velocity difference characteristics of landslide vibration and surge propagation in the spatiotemporal evolution map includes: Identify the near-vertical trajectory formed by the high-frequency landslide vibration component in the spatiotemporal evolution map, and take its initial arrival time as the starting zero point of the landslide water entry event; Identify the inclined trajectory formed by the low-frequency surge component in the spatiotemporal evolution map, and calculate the propagation speed of the surge by least squares fitting based on the slope of the inclined trajectory in the spatiotemporal domain or the time difference with the starting zero point. The criteria for determining a landslide surge event are: simultaneously detecting the near-vertical trajectory and the inclined trajectory, whose propagation speeds differ by more than a preset threshold, within the same spatiotemporal region.

5. The method according to claim 4, characterized in that, The steps for determining a landslide surge event also include: Extract the spatial energy distribution characteristics of the high-frequency landslide vibration component along the fiber optic sensing array; If the spatial energy distribution characteristics satisfy the bidirectional exponential decay model centered on the landslide entry point, then the monitored event is confirmed as a valid landslide surge event. The bidirectional exponential decay model characterizes the physical property that vibration energy decays faster in the near field than in the far field.

6. The method according to claim 1, characterized in that, The determination of the frequency-dependent correction factor based on the frequency response characteristics of fluid-optical fiber interaction includes: Power spectrum analysis is performed on the low-frequency surge component to obtain the power spectral density at different frequencies; based on the ratio of the power spectral density to the preset standard fluid pressure spectral density, the frequency dependence correction factor that varies with frequency is calculated. The fiber frequency domain strain signal and its dependent correction factor were calculated. The relationship between them, the calculation formula satisfies: , in, It is a fiber optic frequency domain strain signal. This is the proportionality coefficient. The dynamic pressure generated by the surge.

7. The method according to claim 6, characterized in that, The step of using landslide deformation correction coefficients determined based on wave train superposition characteristics to perform energy dissipation correction on the wave height retrieved from the hydrodynamic pressure data includes: The number of wave crests and the number of wave train superpositions whose peak amplitudes exceed a preset threshold are counted in the time sequence of the low-frequency surge component. Based on the number of wave train superpositions, a landslide deformation correction coefficient is determined to characterize the degree of landslide fragmentation upon entering the water and the proportion of energy dissipation. ; Introducing landslide deformation correction factor Corrected landslide surge wave height The calculation formula satisfies: , in, For water density, Because of the water depth, The angle between the surge crest line and the fiber axis.

8. A landslide surge wave height inversion system based on distributed optical fiber sensing, characterized in that, include: The signal acquisition module is used to acquire the raw time-series signals collected by the distributed optical fiber sensor array laid on the bottom of the water body; The preprocessing and time-frequency decomposition module is used to denoise and filter the original time-series signal, and to separate the high-frequency landslide vibration component and the low-frequency surge component using time-frequency analysis technology. The event determination and velocity calculation module is used to construct a spatiotemporal evolution map, determine landslide surge events and calculate surge propagation velocity based on the velocity difference characteristics between landslide vibration and surge propagation; The parameter inversion module is used to combine frequency-dependent correction and landslide deformation correction to map the low-frequency surge component into hydrodynamic pressure and correct energy dissipation, and output landslide surge wave height parameters.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 7.

Images