Geological disaster early warning method based on active detection and spectrum response variation

Abstract

This invention relates to the field of geological hazard analysis technology, and particularly to a geological hazard early warning method based on active detection and spectral response variation. The method includes: deploying active signal excitation devices and signal receiving sensors in the monitoring area to apply active detection signals to the geological body and collect response signals; filtering and adaptive time window segmentation of the response signals to obtain stable-period spectral data; establishing a benchmark spectral model through Fourier transform and phase spectrum analysis; periodically transmitting active signals and collecting real-time responses, calculating real-time spectral characteristic parameters, and comparing them with the benchmark spectral model to obtain spectral variability; determining an early warning threshold using a decision tree model based on historical monitoring data; and calculating the geological hazard risk level and outputting early warning information when the real-time spectral variability exceeds the threshold, combined with environmental variables.

Description

Technical Field

[0001] This invention relates to the field of geological disaster analysis technology, and in particular to a geological disaster early warning method based on active detection and spectral response variation. Background Technology

[0002] With the continuous development of transportation, energy, and urban construction in mountainous areas, the frequency of geological disasters is on the rise. Traditional methods for monitoring and early warning of geological disasters mainly rely on macroscopic indicators such as manual inspections, rainfall monitoring, surface displacement observations, or strain gauge measurements. These methods often only reflect changes on the surface or in the shallow layers and are greatly affected by climate, topography, and human factors. They suffer from problems such as long monitoring cycles, low spatial resolution, and delayed response, making it difficult to capture subtle changes in the internal structure of geological bodies in the early stages of disaster incubation, resulting in insufficient accuracy and reliability of early warnings.

[0003] In recent years, with the development of geophysical exploration and signal processing technologies, monitoring the internal state of geological bodies based on active physical exploration has gradually attracted attention. For example, by applying active detection signals such as acoustic waves, elastic waves, or seismic waves, the spectral response characteristics of rock and soil masses can be obtained to analyze their structural integrity and medium changes. In principle, these methods can reflect the elastic characteristics, wave velocity changes, and energy attenuation properties of the internal medium of geological bodies, providing a new technical path for identifying potential structural anomalies. However, existing active detection methods generally use single frequency characteristics, wave velocity, or energy attenuation as diagnostic indicators, lacking multi-dimensional comprehensive quantitative analysis of complete spectral characteristics; at the same time, the dynamic comparison relationship between the stable period spectrum and the real-time spectrum has not been established, making it difficult to accurately characterize the evolution process of the geological body's state, resulting in limited sensitivity in identifying the disaster incubation stage.

[0004] Therefore, there is an urgent need for a technical method that can actively detect and quantify spectral response variations, and achieve real-time monitoring and early warning judgment, so as to improve the accuracy, sensitivity and reliability of geological disaster early warning, thereby providing a scientific basis for disaster prevention and mitigation. Summary of the Invention

[0005] To overcome the above shortcomings, this invention provides a geological disaster early warning method based on active detection and spectral response variation. It aims to improve the shortcomings of existing technologies that rely on passive monitoring, cannot reflect changes in the internal structure of geological bodies in real time, and have delayed and inaccurate early warnings. By actively exciting signals and analyzing spectral responses, it can achieve quantitative identification and early warning of changes in the internal state of geological bodies, thereby significantly improving the reliability and real-time performance of geological disaster monitoring and prevention.

[0006] In a first aspect, the present invention provides the following technical solution: a geological disaster early warning method based on active detection and spectral response variation, comprising:

[0007] Active signal excitation devices and signal receiving sensors are deployed in the geological body monitoring area to apply active detection signals within a preset frequency range to the geological body, collect the stable period response signals of the geological body, and filter and segment the response signals to obtain stable period spectral data.

[0008] Based on the stable period spectral data, the power spectrum, phase spectrum, and energy attenuation parameters are calculated to establish a benchmark spectral model characterizing the stable state of the geological body.

[0009] Periodically transmit active detection signals to geological bodies and collect real-time response signals to extract real-time spectral characteristic parameters;

[0010] The real-time spectral characteristic parameters are compared with the reference spectral model, and the differences between the two in power, phase and energy characteristics are calculated to obtain the spectral variability.

[0011] A decision tree model is established based on historical monitoring data, and the spectral variability is classified and trained to determine the geological disaster early warning threshold.

[0012] When the real-time spectral variability exceeds the warning threshold, the geological disaster risk level and warning information are output.

[0013] Preferably, the active detection signal is a linear sweep frequency signal, a narrowband sine wave signal, or a pulse signal.

[0014] Preferably, the filtering and segmentation steps include:

[0015] The collected geological body stable period response signals are filtered to remove high-frequency noise and low-frequency interference signals, and retain the effective components that are consistent with the frequency band of the active detection signal;

[0016] The filtered signal is segmented according to an adaptive time window to ensure that each segment has relatively stable spectral characteristics.

[0017] Preferably, the adaptive time window acquisition step includes:

[0018] Calculate the frequency gradient within each initial time window for the filtered signal;

[0019] When the frequency gradient is lower than a preset threshold, the time window length is gradually increased to obtain a stable signal segment.

[0020] When the frequency gradient reaches or exceeds a preset threshold, the current time window length is fixed as an adaptive time window, and the processing of the next segment of signal begins.

[0021] Preferably, the steps for establishing the reference spectral model include:

[0022] The power spectrum of the steady-state response signal is calculated using the Fourier transform method.

[0023] The phase spectrum of the steady-state response signal is calculated using phase spectrum analysis.

[0024] The energy decay parameters are calculated by fitting the power spectrum and phase spectrum.

[0025] A reference spectral model is constructed based on the power spectrum, phase spectrum, and energy attenuation parameters.

[0026] Preferably, the step of obtaining the spectral variability includes:

[0027] The power spectrum of the real-time response signal is compared with the power spectrum of the reference spectral model using a spectrum comparison method, and the power spectrum difference is calculated.

[0028] The phase spectrum of the real-time response signal is compared with the phase spectrum of the reference wave spectrum model by using the phase difference analysis method to calculate the phase difference;

[0029] The energy attenuation parameters of the real-time response signal are compared with the energy attenuation parameters of the benchmark spectral model using the energy attenuation comparison method, and the energy attenuation difference is calculated.

[0030] The power spectrum difference, phase difference, and energy attenuation difference are calculated using a weighted combination method to obtain the comprehensive spectral variability.

[0031] Preferably, the step of determining the geological disaster early warning threshold includes:

[0032] Training samples were constructed using historical monitoring data and known geological disaster event data;

[0033] The training samples were classified and trained using a decision tree model to obtain the mapping relationship between spectral variability and geological hazard status.

[0034] The warning threshold is calculated and determined based on the mapping relationship and the preset optimization target.

[0035] Preferably, the steps for outputting geological disaster risk levels and early warning information include:

[0036] The geological body's structural state is determined by comparing the real-time spectral variability with the geological disaster early warning threshold.

[0037] By using risk assessment methods, the judgment results are comprehensively analyzed with environmental variables to calculate the probability and risk level of geological disasters.

[0038] Early warning information is generated based on the risk level, including the geological disaster risk level, the scope of impact, and recommended disaster prevention measures.

[0039] Secondly, the present invention provides the following technical solution: a geological disaster early warning system based on active detection and spectral response variation, used to implement the above-mentioned geological disaster early warning method, the system comprising:

[0040] The stable period spectral data acquisition module is used to deploy active signal excitation devices and signal receiving sensors in the geological body monitoring area, apply active detection signals within a preset frequency range to the geological body, collect the stable period response signals of the geological body, and filter and segment the response signals to obtain stable period spectral data.

[0041] The reference spectral model construction module is used to calculate the power spectrum, phase spectrum and energy attenuation parameters based on the stable period spectral data, and to establish a reference spectral model characterizing the stable state of the geological body.

[0042] The spectral feature extraction module is used to periodically transmit active detection signals to the geological body and collect real-time response signals to extract real-time spectral feature parameters.

[0043] The spectral variability calculation module is used to compare the real-time spectral characteristic parameters with the reference spectral model, calculate the differences between the two in power, phase and energy characteristics, and obtain the spectral variability.

[0044] The disaster early warning threshold determination module is used to establish a decision tree model based on historical monitoring data, classify and train the spectral variability, and determine the geological disaster early warning threshold.

[0045] The risk warning module is used to output the geological disaster risk level and warning information when the real-time spectral variability exceeds the warning threshold.

[0046] The present invention has the following beneficial effects:

[0047] 1. This invention achieves real-time dynamic monitoring of the internal structural state of geological bodies through active detection combined with spectral response analysis. Compared with traditional early warning methods that rely on single indicators such as rainfall, displacement, or deformation, this invention can actively stimulate the response of geological bodies, establish a baseline spectral model for stable periods, and quantify the variability of spectral characteristics in real time, thereby directly reflecting changes in the internal structure of geological bodies. This realizes a shift from passive monitoring to active detection, significantly improving the sensitivity and timeliness of early warning.

[0048] 2. This invention establishes a quantitative mapping relationship between geological hazard status by combining spectral variability with a decision tree model, enabling adaptive determination of early warning thresholds and automatic identification of risk levels. This method utilizes historical monitoring data for model training and combines it with real-time environmental variables for comprehensive evaluation, effectively reducing false alarms and missed alarms, improving the accuracy and reliability of geological hazard early warning, and providing a scientific basis for geological hazard prevention and control. Attached Figure Description

[0049] Figure 1 This is a flowchart of the geological disaster early warning method based on active detection and spectral response variation proposed in this invention;

[0050] Figure 2 This is a structural diagram of the geological disaster early warning system based on active detection and spectral response variation proposed in this invention. Detailed Implementation

[0051] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1

[0053] In the first embodiment of the present invention, the present invention provides a geological disaster early warning method based on active detection and spectral response variation, such as... Figure 1 As shown, it includes the following steps:

[0054] S100: Deploy active signal excitation devices and signal receiving sensors in the geological body monitoring area, apply active detection signals within a preset frequency range to the geological body, collect the stable period response signals of the geological body, and filter and segment the response signals to obtain stable period spectral data.

[0055] Preferably, the active detection signal is a linear sweep frequency signal, a narrowband sine wave signal, or a pulse signal.

[0056] Specifically, based on monitoring requirements and geological characteristics, linear sweep frequency signals, narrowband sinusoidal signals, or pulse signals are selected as active detection signals. Linear sweep frequency signals are suitable for acquiring full-band response information, covering the low-frequency to high-frequency range; narrowband sinusoidal signals are suitable for focusing on the response of specific frequency bands, facilitating the extraction of energy and phase characteristics of the target frequency; pulse signals are suitable for short-time transient response analysis and can be used to detect local structural changes. The generated signal is applied to the geological body to be monitored through an excitation device (such as a vibrator, piezoelectric transducer, or sound source); ensuring that the excitation amplitude is within a safe range while covering the monitoring area, so that the sensor array can collect a complete response signal.

[0057] Preferably, the filtering and segmentation steps include:

[0058] The collected geological body stable period response signals are filtered to remove high-frequency noise and low-frequency interference signals, and retain the effective components that are consistent with the frequency band of the active detection signal;

[0059] The filtered signal is segmented according to an adaptive time window to ensure that each segment has relatively stable spectral characteristics.

[0060] Specifically, the stable-period response signals of the acquired geological bodies are first filtered to remove high-frequency noise and low-frequency interference signals, retaining the effective components consistent with the frequency band of the active probe signals. The filtering process can employ bandpass filters or digital FIR / IIR filtering algorithms, setting low-frequency and high-frequency cutoff frequencies to match the bandwidth of the active probe signals. Signal quality is optimized by adjusting the filter order and filtering parameters to ensure that the filtered signal reflects the true dynamic response of the geological body.

[0061] The filtered signal is segmented according to an adaptive time window. An initial minimum time window length is set, and signal stability is determined by calculating the frequency gradient change amplitude within each time window. If the above indicators are lower than a preset threshold, the time window is gradually extended to obtain a longer stable segment; if the indicators exceed the threshold or reach the maximum window length, the current window length is fixed, and the signal segment is determined to be a relatively stable spectral segment.

[0062] Preferably, the adaptive time window acquisition step includes:

[0063] Calculate the frequency gradient within each initial time window for the filtered signal;

[0064] When the frequency gradient is lower than a preset threshold, the time window length is gradually increased to obtain a stable signal segment.

[0065] When the frequency gradient reaches or exceeds a preset threshold, the current time window length is fixed as an adaptive time window, and the processing of the next segment of signal begins.

[0066] Specifically, the filtered, stable signal is divided into initial minimum time windows, with a minimum window length set to ensure each segment contains sufficient sampling points for spectral analysis. The frequency gradient is calculated for each initial time window to quantify the rate of change of the signal spectrum. One of the following methods can be used: calculating the rate of change of the dominant frequency of the power spectrum over time; calculating the rate of change of the spectral centroid over time; or performing differential or fitting on the frequency components within the sliding window to obtain the gradient value.

[0067] If the calculated frequency gradient is below a preset threshold, it indicates that the signal is relatively stable within that segment. The time window length is then gradually increased to include more continuous and stable signals. If the frequency gradient reaches or exceeds the preset threshold, it indicates a significant change in the signal. The current window length is fixed as an adaptive time window, and this signal segment is marked as a stable segment for subsequent spectral analysis. The window is then slid to the next signal segment, and the frequency gradient calculation and window adjustment steps are repeated until the entire signal sequence has been processed.

[0068] By performing adaptive time window segmentation on the filtered geological body response signal, each segment of the signal maintains relative stability in spectral characteristics such as the main frequency of the power spectrum, the proportion of energy in the frequency band, and the phase spectrum, thereby improving the accuracy and reliability of spectral feature extraction.

[0069] S200: Based on the stable period spectral data, calculate the power spectrum, phase spectrum and energy attenuation parameters, and establish a benchmark spectral model characterizing the stable state of the geological body;

[0070] Preferably, the steps for establishing the reference spectral model include:

[0071] The power spectrum of the steady-state response signal is calculated using the Fourier transform method.

[0072] The phase spectrum of the steady-state response signal is calculated using phase spectrum analysis.

[0073] The energy decay parameters are calculated by fitting the power spectrum and phase spectrum.

[0074] A reference spectral model is constructed based on the power spectrum, phase spectrum, and energy attenuation parameters.

[0075] Specifically, for the stable signal processed by filtering and adaptive time windows, the power spectrum is calculated using Fourier transform to obtain the energy distribution of the signal at each frequency point within each time window. Fast Fourier Transform or Short-Time Fourier Transform can be used to balance time and frequency resolution. To reduce the impact of noise, the power spectrum can be smoothed or weighted using window functions (such as Hanning or Hamming windows).

[0076] For signals within the same stable period, the phase spectrum is calculated using phase spectrum analysis to obtain the phase information of each frequency component. The phase spectrum can be calculated from the real and imaginary parts of the Fourier transform result, or a phase expansion algorithm can be used to eliminate the influence of the 2π jump on subsequent analysis. Multi-segment averaging can be combined during the calculation process to improve the stability and noise resistance of the phase spectrum.

[0077] By combining power spectrum and phase spectrum data, an energy attenuation parameter is calculated using a fitting method to characterize the signal attenuation over time or propagation distance. Exponential attenuation fitting, least squares fitting, or other mathematical models can be selected. Weighted fitting methods can be used during the fitting process to enhance robustness to low signal-to-noise ratio signal segments. Different frequency bands can be fitted separately to obtain frequency-dependent energy attenuation curves.

[0078] By integrating power spectrum, phase spectrum, and energy attenuation parameters, a baseline spectral model for the stable period of a geological body is established. The model can be stored in vector or matrix form, recording the power spectrum, phase spectrum, and corresponding energy attenuation parameters for each time window, facilitating real-time comparison and spectral variability calculation. Further multi-segment averaging or statistical processing can be used to generate an overall baseline model, enhancing the model's robustness and representativeness.

[0079] By calculating the power spectrum and phase spectrum of the stable-period signal after filtering and adaptive time window processing, and fitting the energy attenuation parameters, a benchmark spectral model can be established. This model can accurately characterize the spectral response characteristics of geological bodies in a stable state, ensuring the accurate extraction and robust representation of spectral features. It provides a reliable reference for subsequent spectral variability analysis and geological disaster early warning, thereby improving the accuracy and sensitivity of early warning judgment.

[0080] S300: Periodically emits active detection signals to geological bodies and collects real-time response signals to extract real-time spectral characteristic parameters;

[0081] Specifically, within the geological monitoring area, active detection signals (such as linear sweep signals, narrowband sinusoidal signals, or pulse signals) within a preset frequency range are periodically emitted by deployed active signal excitation devices. Simultaneously, a signal receiving sensor array acquires the geological body's response signals in real time and filters the acquired signals to remove environmental noise. Subsequently, the processed real-time signals are segmented according to a preset or adaptive time window, and the power spectrum and phase spectrum of each signal segment are calculated using Fourier transform or other frequency domain analysis methods. Real-time spectral characteristic parameters are extracted by combining energy attenuation parameters. This process is repeated periodically to continuously obtain the spectral response characteristics of the geological body at different time points, which can be used for subsequent spectral variability analysis and early warning determination.

[0082] S400: Compare the real-time spectral characteristic parameters with the reference spectral model, calculate the differences between the two in power, phase and energy characteristics, and obtain the spectral variability.

[0083] Preferably, the step of obtaining the spectral variability includes:

[0084] The power spectrum of the real-time response signal is compared with the power spectrum of the reference spectral model using a spectrum comparison method, and the power spectrum difference is calculated.

[0085] The phase spectrum of the real-time response signal is compared with the phase spectrum of the reference wave spectrum model by using the phase difference analysis method to calculate the phase difference;

[0086] The energy attenuation parameters of the real-time response signal are compared with the energy attenuation parameters of the benchmark spectral model using the energy attenuation comparison method, and the energy attenuation difference is calculated.

[0087] The power spectrum difference, phase difference, and energy attenuation difference are calculated using a weighted combination method to obtain the comprehensive spectral variability.

[0088] Specifically, the power spectrum of the real-time response signal Power spectrum compared with the reference spectral model To compare the power spectrum differences, the mean square error can be used to calculate the differences.

[0089] ;

[0090] in, For frequency points, For the first One frequency point.

[0091] phase spectrum of real-time signal Phase spectrum with reference Compare and calculate the average phase difference:

[0092] ;

[0093] Real-time signal energy attenuation parameters Compared with the benchmark model By comparison, the difference in energy attenuation is calculated using relative error:

[0094] ;

[0095] The weighted combination of the above three indicators yields the comprehensive spectral variability:

[0096] ;

[0097] in, , and These are the weights for power spectrum difference, phase difference, and energy decay difference, respectively.

[0098] The above methods can quantify the deviation of the real-time response signal from the baseline state, enabling sensitive detection of changes in the internal structure of soil and rock masses. The comprehensive spectral variability can reflect the comprehensive changes in power, phase, and energy attenuation, providing reliable data for subsequent geological disaster early warning threshold determination and improving the accuracy and sensitivity of early warnings.

[0099] S500: Establish a decision tree model based on historical monitoring data, classify and train the spectral variability, and determine the geological disaster early warning threshold;

[0100] Preferably, the step of determining the geological disaster early warning threshold includes:

[0101] Training samples were constructed using historical monitoring data and known geological disaster event data;

[0102] The training samples were classified and trained using a decision tree model to obtain the mapping relationship between spectral variability and geological hazard status.

[0103] The warning threshold is calculated and determined based on the mapping relationship and the preset optimization target.

[0104] Specifically, historical monitoring data is collected, including stable-period spectral characteristics, real-time spectral variability data, and corresponding geological disaster event records (such as landslides, collapses, and subsidence). The data is then organized into a training sample set. ,in, Indicates spectral variability. Indicates the category of geological disaster status (e.g., normal, warning, dangerous). The total number of samples.

[0105] The decision tree classification algorithm is used to train the training samples to establish spectral variability. Geological disaster status The mapping relationship between them. During decision tree training, information gain, Gini index, or other splitting criteria can be used to select an appropriate tree depth and the minimum number of samples per leaf node to prevent overfitting. After training, a decision tree model is obtained. It is used for real-time spectral variability state prediction.

[0106] Based on the mapping relationship obtained from training and the preset optimization objectives (such as maximizing prediction accuracy, minimizing false alarm rate or false alarm rate), the warning threshold corresponding to the spectral variability is determined. The warning threshold can be optimized through cross-validation, ROC curve analysis, or threshold tuning methods to ensure that the expected warning effect is achieved in actual monitoring.

[0107] This method can establish a quantitative mapping between spectral variability and geological hazard status, enabling automated threshold determination. The early warning threshold can dynamically reflect historical data and actual risk levels, improving the accuracy, sensitivity, and reliability of geological hazard early warning.

[0108] S600: When the real-time spectral variability exceeds the warning threshold, output the geological disaster risk level and warning information.

[0109] Preferably, the steps for outputting geological disaster risk levels and early warning information include:

[0110] The geological body's structural state is determined by comparing the real-time spectral variability with the geological disaster early warning threshold.

[0111] By using risk assessment methods, the judgment results are comprehensively analyzed with environmental variables to calculate the probability and risk level of geological disasters.

[0112] Early warning information is generated based on the risk level, including the geological disaster risk level, the scope of impact, and recommended disaster prevention measures.

[0113] Specifically, based on the real-time collected spectral variability, it is compared with a pre-determined geological hazard early warning threshold to determine whether the structural state of the geological body is abnormal. If the spectral variability exceeds the threshold, the geological body is deemed to have potential danger; otherwise, it is considered to be in a normal state. Subsequently, the judgment results are comprehensively analyzed with environmental variables (such as rainfall, groundwater level, topographic slope, and soil type), and a risk assessment method is used to calculate the probability of geological hazard occurrence and classify the risk level, such as low risk, medium risk, or high risk. Early warning information is generated based on the calculated risk level, including the geological hazard risk level, potential impact range, and recommended disaster prevention measures, such as strengthening patrols, early evacuation, or controlling hydrological conditions. This information can be disseminated in real time through monitoring terminals, SMS, mobile applications, or the control center system to achieve rapid response.

[0114] This method combines real-time spectral variability with thresholds and environmental factors to quantify geological hazard risk levels, enabling timely detection and early warning of structural anomalies. It provides intuitive and operable disaster prevention and mitigation basis, improving the accuracy, timeliness, and reliability of early warnings.

[0115] Example 2

[0116] Along a mountainous highway, due to concentrated rainfall and complex geological conditions, there are high-risk areas for landslides and collapses. Traditional monitoring methods mainly rely on manual patrols and rainfall observations, which are difficult to use in a timely and accurate manner to determine changes in the internal structure of geological bodies, resulting in delayed early warnings and the risk of misjudgment.

[0117] To address the aforementioned problems, this invention provides a geological disaster early warning system based on active detection and spectral response variation. The system, as follows... Figure 2 As shown, the system includes modules for acquiring stable-period spectral data, constructing a baseline spectral model, extracting spectral features, calculating spectral variability, determining disaster early warning thresholds, and issuing risk warnings. The specific implementation process is as follows:

[0118] Active signal excitation devices and signal receiving sensors are deployed along key geological sections in the monitoring area; active detection signals within a preset frequency range, such as linear sweep frequency signals, narrowband sinusoidal signals, or pulse signals, are periodically transmitted to the geological body; and the response signals of the geological body are collected in real time to provide raw data for subsequent processing.

[0119] The acquired signal is filtered to remove high-frequency noise and low-frequency interference while retaining the effective frequency band components. The filtered signal is then segmented according to an adaptive time window to ensure that each segment has relatively stable spectral characteristics.

[0120] The power spectrum and phase spectrum of the signal during the stable period are calculated, and the energy attenuation parameters are fitted. The above results are combined to form a reference spectral model, which serves as a reference feature for the geological body under normal conditions.

[0121] Active detection signals are periodically emitted to acquire geological body response signals in real time; power spectrum, phase spectrum and energy attenuation parameters are extracted to calculate the difference with the reference spectrum; the power spectrum difference, phase difference and energy attenuation difference are weighted and combined to obtain the real-time spectrum variability.

[0122] Training samples were constructed using historical monitoring data and known geological disaster event data; a decision tree model was used to classify and train the training samples to establish a mapping relationship between spectral variability and geological disaster status; and early warning thresholds were determined based on the mapping relationship and optimization objectives.

[0123] By comparing real-time spectral variability with early warning thresholds, it is determined whether the geological structure is abnormal; a risk assessment is conducted by integrating environmental variables, the probability of geological disasters is calculated and risk levels are classified; early warning information is generated, including risk level, scope of impact and disaster prevention measures, and is released in real time through monitoring terminals or mobile applications.

[0124] Through the implementation of the aforementioned system, this system can monitor changes in the internal structure of geological bodies in real time, accurately quantify spectral response variations, and combine these variations with historical thresholds and environmental factors for risk assessment, thereby achieving early warning. The application of this system along mountainous highways can significantly improve the accuracy and sensitivity of landslide and collapse warnings, providing a scientific and operational basis for disaster prevention and mitigation.

[0125] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A geological disaster early warning method based on active detection and spectral response variation, characterized in that, include: Active signal excitation devices and signal receiving sensors are deployed in the geological body monitoring area to apply active detection signals within a preset frequency range to the geological body, collect the stable period response signals of the geological body, and filter and segment the response signals to obtain stable period spectral data. Based on the stable period spectral data, the power spectrum, phase spectrum, and energy attenuation parameters are calculated to establish a benchmark spectral model characterizing the stable state of the geological body. Periodically transmit active detection signals to geological bodies and collect real-time response signals to extract real-time spectral characteristic parameters; The real-time spectral characteristic parameters are compared with the reference spectral model, and the differences between the two in power, phase and energy characteristics are calculated to obtain the spectral variability. A decision tree model is established based on historical monitoring data, and the spectral variability is classified and trained to determine the geological disaster early warning threshold. When the real-time spectral variability exceeds the warning threshold, the geological disaster risk level and warning information are output.

2. The geological disaster early warning method based on active detection and spectral response variation according to claim 1, characterized in that, The active detection signal is a linear sweep frequency signal, a narrowband sine wave signal, or a pulse signal.

3. The geological disaster early warning method based on active detection and spectral response variation according to claim 1, characterized in that, The filtering and segmentation steps include: The collected geological body stable period response signals are filtered to remove high-frequency noise and low-frequency interference signals, and retain the effective components that are consistent with the frequency band of the active detection signal; The filtered signal is segmented according to an adaptive time window to ensure that each segment has relatively stable spectral characteristics.

4. The geological disaster early warning method based on active detection and spectral response variation according to claim 3, characterized in that, The steps for obtaining an adaptive time window include: Calculate the frequency gradient within each initial time window for the filtered signal; When the frequency gradient is lower than a preset threshold, the time window length is gradually increased to obtain a stable signal segment. When the frequency gradient reaches or exceeds a preset threshold, the current time window length is fixed as an adaptive time window, and the processing of the next segment of signal begins.

5. The geological disaster early warning method based on active detection and spectral response variation according to claim 1, characterized in that, The steps for establishing a baseline spectral model include: The power spectrum of the steady-state response signal is calculated using the Fourier transform method. The phase spectrum of the steady-state response signal is calculated using phase spectrum analysis. The energy decay parameters are calculated by fitting the power spectrum and phase spectrum. A reference spectral model is constructed based on the power spectrum, phase spectrum, and energy attenuation parameters.

6. The geological disaster early warning method based on active detection and spectral response variation according to claim 1, characterized in that, The steps to obtain spectral variability include: The power spectrum of the real-time response signal is compared with the power spectrum of the reference spectral model using a spectrum comparison method, and the power spectrum difference is calculated. The phase spectrum of the real-time response signal is compared with the phase spectrum of the reference wave spectrum model by using the phase difference analysis method to calculate the phase difference; The energy attenuation parameters of the real-time response signal are compared with the energy attenuation parameters of the benchmark spectral model using the energy attenuation comparison method, and the energy attenuation difference is calculated. The power spectrum difference, phase difference, and energy attenuation difference are calculated using a weighted combination method to obtain the comprehensive spectral variability.

7. The geological disaster early warning method based on active detection and spectral response variation according to claim 1, characterized in that, The steps for determining geological disaster early warning thresholds include: Training samples were constructed using historical monitoring data and known geological disaster event data; The training samples were classified and trained using a decision tree model to obtain the mapping relationship between spectral variability and geological hazard status. The warning threshold is calculated and determined based on the mapping relationship and the preset optimization target.

8. The geological disaster early warning method based on active detection and spectral response variation according to claim 1, characterized in that, The steps for outputting geological disaster risk levels and early warning information include: The geological body's structural state is determined by comparing the real-time spectral variability with the geological disaster early warning threshold. By using risk assessment methods, the judgment results are comprehensively analyzed with environmental variables to calculate the probability and risk level of geological disasters. Early warning information is generated based on the risk level, including the geological disaster risk level, the scope of impact, and recommended disaster prevention measures.

9. A geological disaster early warning system based on active detection and spectral response variation, characterized in that, The system for implementing the geological disaster early warning method according to any one of claims 1-8, the system comprising: The stable period spectral data acquisition module is used to deploy active signal excitation devices and signal receiving sensors in the geological body monitoring area, apply active detection signals within a preset frequency range to the geological body, collect the stable period response signals of the geological body, and filter and segment the response signals to obtain stable period spectral data. The reference spectral model construction module is used to calculate the power spectrum, phase spectrum and energy attenuation parameters based on the stable period spectral data, and to establish a reference spectral model characterizing the stable state of the geological body. The spectral feature extraction module is used to periodically transmit active detection signals to the geological body and collect real-time response signals to extract real-time spectral feature parameters. The spectral variability calculation module is used to compare the real-time spectral characteristic parameters with the reference spectral model, calculate the differences between the two in power, phase and energy characteristics, and obtain the spectral variability. The disaster early warning threshold determination module is used to establish a decision tree model based on historical monitoring data, classify and train the spectral variability, and determine the geological disaster early warning threshold. The risk warning module is used to output the geological disaster risk level and warning information when the real-time spectral variability exceeds the warning threshold.

Images