A method and system for detecting underground diseases based on energy analysis

Abstract

This application discloses a method and system for detecting underground defects based on energy analysis, which involves geological hazards. The method comprises the following steps: obtaining seismic signals from underground structures; dividing the seismic signals into different time windows, each corresponding to a different depth range underground; calculating the energy spectrum of each signal within each time window; calculating the energy change rate P of each signal by comparing the energy spectrum of each signal within each time window with the energy spectrum of each signal in a known dense formation; performing a Hilbert-Huang transform on the obtained seismic signals to obtain an IMF component energy distribution structure diagram and Hilbert marginal spectrum characteristics; and determining the type and location of underground defects based on the energy change rate P, the IMF component energy distribution structure, and the Hilbert marginal spectrum characteristics. Given the limited accuracy of traditional frequency analysis in identifying deep structures and the difficulty in accurately distinguishing between different types of defects, such as cavities and loose bodies, this application accurately identifies different types of underground defects and locates their spatial location and depth range.

Description

Technical Field

[0001] The present application relates to the field of geological disasters, and in particular to a method and system for detecting underground hazards based on energy analysis. Background Art

[0002] With the continuous development of urban construction and underground engineering, underground hazard detection technology is crucial for ensuring project safety. Underground hazards such as cavities and loose bodies, if not promptly discovered and addressed, pose a serious threat to surface buildings and transportation facilities. Therefore, accurately detecting the type and location of underground hazards is a critical step in engineering investigation and safety assessment.

[0003] Subsurface defect detection primarily relies on the acquisition and analysis of seismic signals. Extracting relevant information through signal transformation is a key method used in signal analysis. Currently, subsurface defect detection technology relies primarily on various signal processing methods, such as Fourier transform, short-time Fourier transform (STFT), wavelet transform, and seismic imaging.

[0004] The Fourier transform establishes a connection between the time domain and frequency domain of a signal, forming two ways to observe the signal. This method is convenient, easy to use, and has a clear physical meaning. For stationary signals, the Fourier transform can usually meet the analysis requirements. However, the Fourier transform has obvious defects: (1) Only signals that are absolutely integrable in the time domain can have a Fourier transform; (2) Because the time integration interval is infinite, the time and frequency resolution cannot be met simultaneously; (3) When analyzing seismic signals of underground diseases, in order to mathematically fit the non-stationary waveform of the original data, a large number of high-frequency pseudo-harmonic components will be introduced, resulting in the Fourier spectrum underestimating the low-frequency energy, affecting the detection accuracy.

[0005] The short-time Fourier transform (STFT) is commonly used to analyze nonstationary signals. Its characteristic is that it can obtain the frequency components of the overall signal in the local time domain through a plane composed of two coordinates: the time axis and the frequency axis, allowing the distribution and arrangement of the individual frequency bands of the entire signal at a local time. However, this method must assume that the signal is piecewise stationary. Once the window function is determined, its time and frequency resolutions remain constant throughout the signal's time-frequency range, resulting in a lack of flexibility and difficulty adapting to the complex and changing characteristics of underground media.

[0006] The wavelet transform extracts and analyzes the signal by applying a variable-scale sliding window to it. The width of the time and frequency windows is variable, resulting in excellent localization properties in both the time and frequency domains. Different locations on the time-frequency plane have different resolutions, effectively resolving the conflict between time resolution and frequency resolution. However, the wavelet transform is subject to many constraints in practical applications, such as algorithm complexity and stationarity constraints, and still has limitations when processing nonlinear and non-stationary signals used in underground disease detection.

[0007] Seismic imaging (also known as high-density seismic exploration and multi-wave seismic exploration) is based on the optimal offset technique used in reflection wave exploration. This method can utilize multiple seismic waves as effective waves for exploration, or it can select only a specific wave as the effective wave, depending on the exploration objective. Seismic imaging is effective when the exploration objective is relatively simple, focusing on lateral geological changes. However, determining the optimal offset is difficult when exploring multiple target layers. Because each trace uses the same offset, temporal variations in the seismic record primarily reflect the underlying geological structure, facilitating data interpretation. However, traditional seismic imaging reflection wave analysis for determining subsurface structure is typically manual and lacks automation and intelligence. Furthermore, the exploration process often requires frequent changes in the excitation conditions to obtain different responses, resulting in complex operations and poor consistency.

[0008] Existing methods for detecting underground defects generally rely on frequency analysis rather than energy analysis, limiting their ability to identify deep structural changes. Furthermore, existing methods are inadequate for processing nonlinear and non-stationary seismic signals, making it difficult to accurately identify and locate various underground defects, particularly in complex geological conditions, such as distinguishing between cavities and loose bodies. Summary of the Invention

[0009] In response to the problems of limited accuracy of traditional frequency analysis in identifying deep structures and difficulty in accurately distinguishing different types of defects such as cavities and loose bodies, the present application provides an underground defect detection method and system based on energy analysis. This method is based on an underground defect detection method that combines energy analysis of each channel with time window depth mapping under constant excitation energy conditions. It can effectively process nonlinear and non-stationary seismic signals while maintaining consistent excitation conditions through energy change rate calculation and Hilbert marginal spectrum feature analysis, accurately identify different types of underground defects, and precisely locate their spatial position and depth range.

[0010] One aspect of the present application provides an underground defect detection method based on energy analysis, comprising: acquiring seismic signals of underground structures; dividing the seismic signals into different time windows, each time window corresponding to a different depth range underground; calculating the energy spectrum of each signal in each time window; calculating the energy change rate P of each signal by comparing the energy spectrum of each signal in each time window with the energy spectrum of each signal in a known dense formation; performing a Hilbert-Huang transform on the acquired seismic signals to obtain Hilbert marginal spectrum characteristics; and determining the type and location of the underground defect based on the energy change rate P and the Hilbert marginal spectrum characteristics.

[0011] Furthermore, obtaining seismic signals of underground structures includes: collecting seismic reflection wave signals as seismic signals through a detector system; wherein the detector system includes: arranging multiple detectors according to preset equal intervals to form a detector array so that the detector interval is constant; controlling the height and weight of each drop hammer to be the same so that each drop hammer excites the same energy; and through positioning marks, making the distance between the excitation point and the first detector in each measurement constant so that the small offset distance is the same.

[0012] Furthermore, the seismic signal is divided into different time windows, each time window corresponding to a different depth range underground, including: obtaining a time-depth conversion relationship based on the average propagation speed of the seismic signal in the underground medium; dividing the seismic signal into multiple time windows according to a preset time interval; establishing a mapping relationship between each time window and the corresponding depth range based on the time-depth conversion relationship, and the mapping relationship is used for hierarchical identification of underground structures and disease types; wherein, stratification refers to underground structural layers of different depth ranges, each depth layer contains multiple signals, and each signal corresponds to data collected by detectors at different positions.

[0013] Furthermore, the energy spectrum of each signal in each time window is calculated, including: preprocessing the seismic signal in each time window; calculating the energy spectrum of each signal in each time window based on the preprocessed seismic signal; dividing the test section according to preset intervals, and calculating the average energy value of each divided section.

[0014] Furthermore, the energy change rate of each signal is calculated, including: selecting a known dense layer as a reference benchmark; calculating the energy change rate P of each signal relative to the reference benchmark based on the average energy value of each divided section; and dividing the density state of the underground structure of each divided section according to the energy change rate.

[0015] Specifically, first, from the perspective of wave theory, wave propagation in dense homogeneous media follows the classical wave equation, with energy primarily decaying geometrically. For waves of the same frequency, the decay rate is stable and predictable. This stability provides a theoretical basis for establishing a "standard reference system." Second, from the perspective of energy transfer, the energy-frequency distribution curve in dense layers exhibits stable morphological characteristics, typically manifested as energy peaks in specific frequency bands and expected energy decay slopes. This characteristic makes dense layers an ideal "baseline" for distinguishing the states of other media. Third, from the perspective of statistical robustness, the energy characteristics of dense layers are highly consistent and repeatable across space, reducing the impact of random errors on judgment. By using the dense layer as a reference, the system effectively establishes a "relative energy measurement" mechanism, effectively eliminating the influence of environmental variables and equipment differences.

[0016] Furthermore, the energy change rate P of each signal relative to the reference standard is calculated using the following formula: P = (Em - Er) / Er, where P is the energy change rate, Em is the average energy value of the signal to be measured in the corresponding divided segment, and Er is the average energy value of the corresponding reference standard dense channel signal in the same divided segment.

[0017] In particular, the energy change rate P = (Em - Er) / Er. This formula essentially quantifies the degree of wave energy modulation by the test medium. A positive rate of change indicates that the test medium attenuates the wave energy less or reflects it more strongly than the reference medium; a negative rate of change indicates that the test medium absorbs or scatters more wave energy. Furthermore, it eliminates the influence of differences in the original absolute value of the energy, making measurement results comparable across different geological backgrounds and excitation conditions. This processing approach is similar to the concept of adaptive filtering, but its goal is to extract energy modulation characteristics rather than the signal itself.

[0018] Furthermore, the density state of the underground structure in each divided section is divided according to the energy change rate, including: when the energy change rate is less than the threshold value ε1, it is judged to be moderate to severe loose disease, and the density state is low density; when the energy change rate is greater than or equal to the threshold value ε1 and less than the threshold value ε2, it is judged to have abnormal characteristics of loose disease, and the density state is medium density; when the energy change rate is greater than or equal to the threshold value ε2 and less than the threshold value ε3, it is judged to be a normal section, and the density state is dense; when the energy change rate is greater than or equal to the threshold value ε3 and less than the threshold value ε4, it is judged to be a small cavity or loose disease, and the density state is medium density; when the energy change rate is greater than the threshold value ε4, it is judged to be a large cavity or severe loose disease, and the density state is low density.

[0019] Specifically, first, from the perspective of wave propagation in multiphase media, this segmented determination method accurately captures the bidirectional variation of "porosity-density-voidity." Traditional methods often use a one-way threshold determination (e.g., a value below a certain threshold is considered abnormal), which fails to distinguish between porous media and voids, two anomalies with distinct physical properties. Second, from the perspective of energy conversion mechanisms, the threshold divisions precisely correspond to different physical processes: a significant decrease in energy (P < ε1) corresponds to wave energy absorption by the porous medium; a slight decrease (ε1 ≤ P < ε2) corresponds to slight porosity; a normal range (ε2 ≤ P < ε3) corresponds to a dense layer; a slight increase (ε3 ≤ P < ε4) corresponds to boundary reflection from small voids; and a significant increase (P ≥ ε4) corresponds to strong reflection from large voids. Third, from the perspective of the mapping relationship between signal characteristics and physical structure, this nonlinear segmented determination method effectively overcomes the limitations of traditional linear mapping methods. The relationship between subsurface structure and signal energy is inherently nonlinear, and simple linear mapping cannot accurately reflect complex subsurface damage conditions.

[0020] ε1 (usually -50% to -40%): represents the upper limit of the typical absorption of wave energy by loose media. A value smaller than this value indicates that the medium has significant energy attenuation characteristics; ε2 (usually -40% to -20%): represents the lower limit of energy fluctuation allowed by measurement errors and natural geological variations; ε3 (usually 10% to +20%): represents the upper limit of energy fluctuation allowed by measurement errors and natural geological variations; ε4 (usually 20% to +40%): represents the lower limit of energy amplification caused by typical cavity boundary reflection.

[0021] Furthermore, empirical mode decomposition and Hilbert-Huang transform are performed on the acquired seismic signals to obtain the energy structure distribution characteristics and Hilbert marginal spectrum characteristics of the IMF components, including: performing EMD decomposition on the acquired seismic signals to obtain IMF components, obtaining an energy structure distribution diagram of the IMF components of each data, reflecting the energy proportion of IMF components of different frequencies. When there are diseases in the ground, the energy structure distribution of the IMF components will change significantly.

[0022] Perform Hilbert-Huang transform on the signal to obtain the Hilbert marginal spectrum features. Extract the maximum energy, the instantaneous frequency corresponding to the maximum energy, and the decay rate after the energy reaches the maximum value from the Hilbert marginal spectrum as the Hilbert marginal spectrum features:

[0023] (1) Maximum energy: In a normal dense formation, the maximum energy of the marginal spectrum is much greater than the maximum energy of the formation containing loose bodies.

[0024] (2) Maximum energy instantaneous frequency: In normal formations, energy reaches its maximum value at low frequency, while in some formations containing loose bodies, the maximum energy instantaneous frequency will increase significantly.

[0025] (3) Energy decay rate: In normal formations, as the instantaneous frequency increases, the energy reaches its maximum value at low frequency and then decays rapidly. In formations containing loose bodies, after the energy reaches its maximum value, the decay rate is relatively slow.

[0026] Furthermore, the type and location of underground diseases are determined based on the energy change rate P, the IMF component energy structure diagram and the Hilbert marginal spectrum characteristics, including: when the energy change rate P is greater than or equal to the threshold Δ1, the corresponding disease type is determined to be a cavity; when the energy change rate P is less than the threshold Δ2, and the Hilbert marginal spectrum characteristics and the IMF component energy distribution characteristics meet any of the following conditions, the disease type is determined to be a loose body: (1) the decrease in the maximum energy value in the Hilbert marginal spectrum is greater than the threshold β1; (2) the increase in the instantaneous frequency corresponding to the maximum energy value in the Hilbert marginal spectrum is greater than the threshold β2; (3) the decrease in the attenuation rate after the energy reaches the maximum value in the Hilbert marginal spectrum is greater than the threshold β3; (4) the similarity of the IMF component energy structure is less than the threshold β4;

[0027] According to the mapping relationship between each time window and the corresponding depth range in step S2 and the division result of the underground structure density state in step S4, the location and depth range of the underground disease are determined.

[0028] In particular, the energy change rate P reflects the modulation effect of different media on seismic waves from the perspective of wave propagation: Cavity: When seismic waves encounter a cavity (solid-air interface), due to the huge difference in acoustic impedance (the acoustic impedance of solid is about 10 of that of air), .... 4 times), the reflection coefficient R ≈ -1, meaning that almost all the wave energy is reflected. In multi-channel recordings, this strong reflection results in superposition and enhancement between adjacent channels, resulting in a significantly positive rate of change of energy P. In the case of loose bodies, when waves propagate through loose media, energy loss arises from three physical mechanisms: ① heat loss due to friction at the interface between particles (internal friction attenuation); ② viscous damping of the pore fluid (viscous attenuation); and ③ scattering due to the irregular distribution of particles and pores (scattering attenuation). These three mechanisms work together to make the rate of change of energy P significantly negative.

[0029] In addition, Hilbert marginal spectrum analysis is based on the Hilbert-Huang transform (HHT) and extracts the intrinsic oscillation modes of the signal through empirical mode decomposition (EMD), avoiding the limitations of the preset basis functions of the Fourier transform and is more suitable for nonlinear and non-stationary signals such as earthquakes; it can accurately characterize the changes in instantaneous frequency and instantaneous amplitude, and capture the subtle modulation characteristics of the medium on the signal; it can effectively characterize the nonlinear modulation characteristics of different media on fluctuations, which is difficult to achieve with traditional linear analysis methods.

[0030] Finally, the energy change rate P ≥ Δ1 enables precise identification of voids in principle: the Δ1 threshold (typically set to 150% to 200%) corresponds to the energy ratio of the reflected wave to the incident wave, caused by the acoustic impedance difference at the void boundary. Void reflections exhibit three typical characteristics on the Hilbert marginal spectrum: broadband energy enhancement, reflecting the high reflectivity of the void boundary across all frequency bands; phase jumps, reflecting the phase characteristics of the reflected wave relative to the direct wave; and time-frequency clustering, reflecting the localized and sudden nature of the reflection event. This dual-conditional discrimination mechanism effectively avoids misjudgments that could result from energy changes alone (e.g., increased density of hard interlayers can also lead to energy enhancement).

[0031] The combined discrimination strategy of energy change rate P<Δ2 + Hilbert marginal spectrum, which satisfies any of the three characteristics, achieves highly adaptive recognition of loose bodies:

[0032] The Δ2 threshold (typically set between -30% and -50%) corresponds to the combined effect of the three main attenuation mechanisms in porous media: scattering, internal friction, and viscous damping. According to wave attenuation theory, the attenuation coefficient α is proportional to the frequency to the power of n, where n depends on the dominant attenuation mechanism: scattering attenuation is n≈4 (in the Rayleigh scattering regime) or n≈1 (in the geometric scattering regime); internal friction attenuation is n≈1; and viscous attenuation is n≈2.

[0033] The three characteristics of the Hilbert marginal spectrum correspond to the different physical properties of loose bodies:

[0034] Energy Maximum Reduction (Threshold β1): This reflects the overall attenuation effect of the porous medium on the wave energy. β1 (typically 20% to 40%) is set based on the typical attenuation of the total signal energy by the porous medium.

[0035] Increased maximum energy frequency (threshold β2): This reflects the dispersion effect and selective attenuation characteristics of porous media. Loose media attenuate high-frequency components more strongly, shifting the spectral center of gravity toward lower frequencies and increasing the frequency corresponding to the maximum energy point. β2 (typically 15% to 30%) is set based on the morphology of the dispersion curve.

[0036] Reduced energy decay rate (threshold β3): This reflects the modulation of the energy spectrum by the porous medium. Multiple scattering in the porous medium causes energy redistribution in the frequency domain, manifesting as slower energy decay at high frequencies. β3 (typically 25% to 45%) is set based on the typical intensity of the spectral tailing effect.

[0037] The characteristics of the IMF component energy structure diagram correspond to the physical properties of the loose body:

[0038] A decrease in the similarity of the IMF component energy structure (threshold β4): This reflects a significant change in the contribution of different frequency components within the bulk to the total signal energy. (β4 is typically 35% to 45%). This setting is based on the change in the contribution of each IMF component energy to the total signal energy within that channel.

[0039] The adoption of the judgment strategy of "meeting any condition" is the key innovation of this scheme, which reflects the adaptation to the diversity of loose bodies: loose bodies of different causes and structures may show different performances in these three characteristics, and the flexible judgment logic greatly improves the adaptability of identification.

[0040] Another aspect of the present application provides an underground defect detection system based on energy analysis, comprising: a signal acquisition module, which collects seismic reflection wave signals through a detector system, wherein the detector system includes a detector array formed by multiple detectors arranged at preset equal intervals; a time window module, which divides the seismic signal into different time windows and establishes a mapping relationship between each time window and a corresponding depth range based on a time-depth conversion relationship; an energy spectrum module, which preprocesses the seismic signal within each time window, calculates the energy spectrum of each signal within each time window, and calculates the average energy value of each divided segment; an energy change rate module, which calculates the energy change rate P of each signal by comparing the energy spectrum of each signal within each time window with the energy spectrum of each signal of a known dense formation; a Hilbert-Huang module, which performs empirical mode decomposition and Hilbert-Huang transform on the acquired seismic signal to obtain Hilbert marginal spectrum characteristics; and a defect judgment module, which judges the type and location of the underground defect based on the energy change rate P and the Hilbert marginal spectrum characteristics, and determines the depth range of the underground defect based on the mapping relationship between the time window and the depth range.

[0041] Compared with the existing technology, the advantages of this application are:

[0042] When detecting different types of underground diseases and their locations, existing technologies generally use frequency analysis methods based on Fourier transform or seismic imaging methods that require frequent changes in excitation conditions. However, they have defects such as limited ability to process nonlinear and non-stationary seismic signals, low accuracy in identifying deep structures, and difficulty in accurately distinguishing different types of diseases. This application combines energy spectrum analysis under constant excitation energy conditions with time window depth mapping and applies Hilbert-Huang transform processing technology to achieve effective processing of nonlinear and non-stationary seismic signals, significantly improving the accuracy of identifying underground disease types and depth positioning, and especially enhancing the ability to distinguish between cavities and loose bodies. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The present application will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, the same numbers represent the same structures, wherein:

[0044] Figure 1 is an exemplary flow chart of a method for detecting underground diseases based on energy analysis according to some embodiments of the present application;

[0045] Figure 2 is a diagram of an experimental location according to some embodiments of the present application;

[0046] Figure 3 This is a schematic diagram of experimental measurement line layout according to some embodiments of the present application;

[0047] Figure 4 is a schematic diagram of energy curves of different formations according to some embodiments of the present application;

[0048] Figure 5 It is a waveform and IMF energy distribution structure diagram of loose formation and dense formation shown in some embodiments of the present application;

[0049] Figure 6 is a schematic diagram of a marginal spectrum of a dense formation according to some embodiments of the present application;

[0050] Figure 7 is a schematic diagram of a marginal spectrum of a loose formation according to some embodiments of the present application;

[0051] Figure 8 is a schematic diagram of the terrain structure of a test section according to some embodiments of the present application;

[0052] Figure 9 This is a comparison diagram of the energy of two tests on a test section according to some embodiments of the present application. DETAILED DESCRIPTION

[0053] The method and system provided in the embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0054] This application is mainly used in urban engineering surveys, underground road disease detection and other fields. Under the premise of ensuring the same small offset and the same excitation energy for each drop hammer, the energy spectrum change rate of the signal in different time windows is analyzed, and the Hilbert-Huang transform analysis of the signal's IMF component energy characteristics and the signal's marginal spectrum characteristics are analyzed to determine the soil density within different depth ranges.

[0055] Experiments have confirmed that frequency is not very useful for distinguishing dense strata from cavities. Using seismic images to observe changes in signal energy and event morphology is more effective. For detecting loose bodies, seismic images show a significant decrease in signal amplitude (energy) and event dislocation, so signal energy is the primary method of analysis.

[0056] The IMF energy structure distribution diagram and Hilbert marginal spectrum obtained by HHT can reflect the distribution of energy at different frequencies. When seismic waves propagate underground, if damage occurs in a certain area, energy transmission will be affected. The energy value corresponding to each instantaneous frequency will inevitably show a certain variation pattern, causing fluctuations and mutations in the IMF energy distribution structure and marginal spectrum. By comparing each marginal spectrum, we can identify the location of suspected damage. This method is particularly effective for identifying loose bodies.

[0057] It was finally determined that signal energy would be mainly used for underground structure analysis, and HHT transformation would be used to assist in loose body analysis and judgment.

[0058] like Figure 1 As shown, the seismic signal of the underground structure is obtained; the seismic signal is divided into different time windows, each time window corresponds to a different depth range underground; the energy spectrum of each signal in each time window is calculated; the energy change rate P of each signal is calculated by comparing the energy spectrum of each signal in each time window with the energy spectrum of each signal of a known dense formation; the Hilbert-Huang transform is performed on the obtained seismic signal to obtain the Hilbert marginal spectrum characteristics; the type and location of the underground disease are determined based on the energy change rate P and the Hilbert marginal spectrum characteristics.

[0059] Specifically, obtaining seismic signals from underground structures includes: collecting seismic reflection wave signals as seismic signals through a detector system; wherein the detector system includes: arranging multiple detectors according to preset equal intervals to form a detector array so that the detector interval is constant; this uniform sampling enables the precise calculation of the energy transfer and attenuation laws in space in subsequent signal processing.

[0060] The height and weight of each hammer drop are controlled to be consistent, ensuring that each drop generates the same energy; this contrasts sharply with conventional methods that require frequent changes in excitation conditions. This constant excitation energy ensures a consistent normalization baseline for all measured data, accurately reflecting energy variations caused solely by differences in the subsurface medium. This avoids multivariate interference and improves the ability to handle nonlinear and nonstationary signals.

[0061] By positioning the markers, the distance between the excitation point and the first detector is kept constant throughout each measurement, ensuring consistent small offsets. This provides a stable reference baseline for establishing accurate time-depth relationships, effectively addressing the accuracy issues of traditional methods in identifying deep structures.

[0062] This application uses the control variable method to take the underground medium characteristics as the only variable, so that the system's response to underground diseases depends only on the underground structure itself, thereby overcoming the limitations of traditional frequency analysis methods in processing nonlinear and non-stationary signals.

[0063] Seismic signals are divided into distinct time windows, each corresponding to a different depth range. A time-depth conversion relationship is derived based on the average propagation velocity of the seismic signal in the subsurface medium. This conversion relationship is then used to establish a precise time-space mapping. This conversion relationship is based on the physical laws of wave propagation. However, unlike traditional methods, this approach not only considers a single average velocity but also implicitly adapts to the stratified characteristics of the subsurface medium, laying the foundation for precise positioning in complex geological conditions.

[0064] Seismic signals are divided into multiple time windows at preset intervals. While traditional Fourier transform methods lose local features when processing the entire signal, this approach achieves segmented, quasi-stationary processing of non-stationary signals through time windowing. The signal characteristics within each window are more stable, facilitating subsequent detailed analysis. This processing method essentially localizes the signal in time and frequency, but avoids the limitations of the fixed window of the short-time Fourier transform.

[0065] Based on the time-depth conversion relationship, a mapping relationship is established between each time window and the corresponding depth range. This mapping relationship is used for hierarchical identification of underground structures and disease types. Hierarchies refer to underground structural layers with different depth ranges. Each depth layer contains multiple signals, each corresponding to data collected by a detector at a different location. Through this mapping, the system can directly locate anomalies in the actual geological space, rather than simply identifying features in the abstract signal space. In particular, this mapping mechanism introduces the concept of "hierarchical identification," which decomposes three-dimensional spatial problems into a manageable combination of two-dimensional problems. Specifically, the lateral changes of multiple signals are analyzed within each depth layer, significantly reducing the computational complexity under complex geological conditions.

[0066] This design allows the application to focus on analyzing energy changes caused solely by differences in underground media, providing a theoretical basis for accurately distinguishing between different types of damage, such as cavities and loose bodies. Traditional methods often struggle to distinguish between different types of damage due to overlapping signal interference. However, this approach, through depth stratification and lateral comparison, effectively isolates the mutual interference of signals at different depths, significantly improving identification accuracy.

[0067] Calculating the energy spectrum of each signal in each time window includes: preprocessing the seismic signals in each time window; calculating the energy spectrum of each signal in each time window based on the preprocessed seismic signals; dividing the test section according to preset intervals, and calculating the average energy value of each divided section.

[0068] In particular, this solution uses energy spectrum analysis, rather than traditional frequency spectrum analysis, as its core identification method. The energy spectrum directly reflects the energy transfer and absorption characteristics of seismic waves in different media and is more sensitive to changes in the physical properties of underground media. Suspected damage points are identified by determining the rate of change of energy across each channel within the same time window. This identification method based on relative spatial changes is more adaptable than traditional absolute threshold methods and can effectively meet detection needs in diverse geological settings.

[0069] More specifically, while cavities are not obvious in frequency signatures, this approach reveals that cavities manifest as bright reflections with strong amplitude on seismic images, significantly increasing their energy. This energy enhancement is due to the strong reflection of seismic waves at the air-rock interface, a characteristic that is difficult to capture effectively using traditional frequency analysis methods.

[0070] On the other hand, the dominant frequency of loose bodies is lower than that of hard formations. HHT decomposition reveals that the energy structure of the intrinsic mode function (IMF) components of loose bodies differs significantly from that of dense formations. In the HHT marginal spectrum, loose bodies exhibit a unique "three decreases and one slowdown" characteristic: a significant decrease in maximum energy value, a delay in maximum energy frequency, and a decrease in energy decay rate. Furthermore, this scheme identifies the "pulling down of event axes and reducing energy" phenomenon produced by loose bodies on seismic profiles, a finding that provides additional discriminant evidence for loose body identification.

[0071] In summary, this solution not only identifies and locates underground damage but, more importantly, can distinguish between different types of underground damage, particularly cavities (energy increases) and loose bodies (energy attenuation), two of the most common and easily confused types. This ability stems from a deep understanding of the physical mechanisms of seismic wave propagation in different media and the application of innovative signal processing methods.

[0072] In this embodiment, for the seismic image amplitude energy, in the same time window, the signal energy of each channel is compared with the energy of the known dense formation channel. Through multiple experiments, we set the energy change rate and the corresponding underground density and hidden danger development characteristics, as shown in Table 1.

[0073] Table 1 Classification criteria for density state

[0074] Energy change rate (P) Development characteristics of underground hidden dangers dense state P<-60% Moderate to severe loosening disease Low density -60≤P<-30% Possesses abnormal characteristics of loose disease Medium density -30≤P<20% Normal road section dense 20≤P<40% Small voids and loose diseases Medium density P≥40% Large cavities and severe looseness Low density

[0075] The experimental site is the loose backfill area of the fish pond in the yard of a construction engineering company. The backfill area is about 10 meters long, 2 meters wide and 1.5 meters deep. The measuring line extends from south to north to the cement hard road surface (dense layer). The experimental site is as follows Figure 2 The experimental line layout diagram is shown in Figure 3 As shown in the figure, the energy curves of different formations are obtained, such as Figure 4 The energy comparison of the dense hard road surface, the soil layer outside the backfill area, and the soil layer inside the backfill area is shown in Table 2:

[0076] Table 2 Energy and energy change rate of different formations

[0077] Place Compacted road surface Land outside fish pond Land within fish ponds Energy mean 51.08155152 35.11079002 27.39678127 Energy change rate / -31.27% -46.37%

[0078] According to the energy change rate setting standards corresponding to different degrees of diseases, the backfill area was determined to have abnormal characteristics of loose diseases, but it was not as loose as moderate or severe, which basically met the on-site soil conditions.

[0079] We analyze the percentage of the energy of each IMF component in the total energy, observe the distribution of the IMF component energy, and express the percentage of each component in different colors. The higher the percentage, the longer the corresponding color band.

[0080] The waveform and IMF energy distribution structure of loose strata and dense strata are as follows Figure 5 As shown in the figure, it can be observed that in the IMF energy spectrum, the energy proportion structure of the IMF components in the loose part (left side) and the dense layer (right side) also has obvious characteristic differences.

[0081] This application compares the marginal spectra of loose and dense channels to obtain the marginal spectrum of dense layers, such as Figure 6 As shown, the marginal spectrum of loose strata, such as Figure 7 As shown in Figure 2, by comparing the normal dense stratum with the stratum containing loose bodies, we can notice four obvious parameter changes:

[0082] (1) Maximum energy: In a normal dense formation, the maximum energy is much greater than the maximum energy of a formation containing loose bodies, and the decrease in the maximum energy in a loose formation is greater than the threshold β1;

[0083] (2) Maximum energy instantaneous frequency: In normal formations, energy reaches its maximum value at low frequency, while in some formations containing loose bodies, the maximum energy instantaneous frequency will increase significantly, and the increase is greater than the threshold β2;

[0084] (3) Energy decay rate: In normal formations, as the instantaneous frequency increases, the energy reaches its maximum value at low frequencies and then decays rapidly. In formations containing loose bodies, after the energy reaches its maximum value, the decay rate is relatively slow. The decay rate decreases by more than the threshold value β3.

[0085] (4) The similarity of the IMF component energy structure is less than the threshold β4;

[0086] This application monitors a road section to be tested in a certain area. The monitoring section is a bicycle lane from north to south in the road section to be tested. From north to south, about 20 meters south of the red iron cage, there is a newly excavated subway station under the road surface with a top depth of 13 meters. The underground structure diagram is as follows Figure 8 As shown, data were collected on the road sections to be measured on August 16, 2024 and November 11, 2024, respectively. Data were collected on the same road section according to the coordinates. During the collection process, the instrument moved forward at a constant speed.

[0087] Within 3 meters underground, the collected data was de-badged and band-pass filtered, and the energy spectra of the two data were analyzed. The changes in the energy spectra were compared and analyzed, and the energy diagrams of the two tests were obtained as follows: Figure 9 As shown, the overall quality of the first phase of data is good, the overall stratum of the monitored section is relatively dense, the stratum density starting from the north side (left) is slightly lower than that of the south (right), and the data energy on the south side is slightly higher than that on the north side. When the equipment passes through the manhole cover (cavity), the phase axis will be disconnected and the amplitude will suddenly increase.

[0088] The overall quality of the second phase of data is good. The overall stratum of the monitored section is relatively dense. The stratum density starting from the north side (left) is slightly lower than that of the south (right) side. The data energy on the south side is slightly higher than that on the north side. When the equipment passes through the manhole cover (cavity), there will be obvious misalignment of the phase axis and a sudden increase in amplitude, which is consistent with the first phase of data.

[0089] When the structural state of underground soil changes, especially when it is disturbed and loose, energy dissipation will occur during the propagation of seismic waves, and the energy spectrum will decrease. As the loose state becomes stronger, the reduction rate will increase. On the other hand, when the soil is seriously loose or even forms voids, the wave group frame will be significantly enhanced during the propagation of seismic waves, strong reflection will occur, and the energy spectrum will be enhanced again. Therefore, it is proposed to use the rate of change of energy spectrum as a classification of soil density state to reflect the impact of soil structural disturbance on the formation and development of underground disease bodies, so as to facilitate the timely discovery of potential underground disease hazards that may cause collapse.

[0090] At the same time, the waveform data collected and processed by the broadband vibration monitoring system studied in this project can also identify the development of underground disease hazards through data collection and analysis. If the development characteristics of underground disease bodies are identified through data collection and analysis, it reflects that underground disease hazards have been formed, and it is necessary to carry out planned detection. Therefore, the identification of the development characteristics of underground disease bodies is used as a standard for monitoring the compaction state.

[0091] To this end, a method for identifying changes in soil compaction state based on the degree of underground disease development and energy spectrum change rate indicators was established and classified.

[0092] We performed average energy analysis on a 20-meter section, using windows of 2-4 meters as a window. We analyzed the energy changes within each window to determine changes in the compaction state. The energy spectrum changes between the two tests are shown in Table 3.

[0093] Table 3 Changes in energy spectrum between two tests

[0094] Segment number Starting coordinates End coordinates First Energy Second energy Energy change rate Soil density 1 X=4418302.9Y=454130.75 X=4418300.2Y=454130.76 22.02 22.23 0.95% dense 2 X=4418300.2Y=454130.76 X=4418296.33Y=454130.76 35.98 34.07 -5.31% dense 3 X=4418296.33Y=454130.75 X=4418292.00Y=454130.75 26.00 25.38 -2.38% dense 4 X=4418292.00Y=454130.78 X=4418288.1Y=454130.78 17.84 14.56 -18.39 dense 5 X=4418288.1Y=454130.79 X=4418285.32Y=454130.79 40.45 40.05 -0.99% dense 6 X=4418285.32Y=454130.80 X=4418283.33Y=454130.80 26.74 27.93 4.45% dense

[0095] The road section under test is generally in good condition. The pavement above the northern part of the underground subway station is slightly looser than that to the south. Overall, the ground is relatively compact, and the difference between the two monitoring data is small, indicating no significant change in the ground compactness beneath the second lane of the north-south section under test.

[0096] The above schematically describes the invention of the present application and its implementation methods. This description is not restrictive. Without departing from the spirit or basic features of the present application, the present application can be implemented in other specific forms. What is shown in the drawings is only one of the implementation methods of the invention of the present application. The actual structure is not limited to this. Any figure mark in the claims should not limit the claims involved. Therefore, if a person of ordinary skill in the art is inspired by it, without departing from the purpose of the invention, a structural method and embodiment similar to the technical solution are designed without creativity, which should fall within the scope of protection of the present application. In addition, the word "including" does not exclude other elements or steps, and the word "one" before an element does not exclude the inclusion of "multiple" elements. The multiple elements stated in the product claim can also be implemented by one element through software or hardware. Words such as first and second are used to indicate names and do not indicate any specific order.

Claims

1. A method for detecting underground diseases based on energy analysis, characterized in that: include: Acquire seismic signals from underground structures; Divide the seismic signal into different time windows, each time window corresponds to a different depth range underground; Calculate the energy spectrum of each signal in each time window; By comparing the energy spectrum of each signal in each time window with the energy spectrum of each signal in the known dense layer, the energy change rate P of each signal is calculated; Perform Hilbert-Huang transform on the acquired seismic signal to obtain the Hilbert marginal spectrum characteristics; Determine the type and location of underground diseases based on the energy change rate P and Hilbert marginal spectrum characteristics; Acquire seismic signals of underground structures, including: The seismic reflection wave signal is collected by the geophone system as a seismic signal; The detector system includes: arranging multiple detectors at preset equal intervals to form a detector array so that the detector spacing is constant; controlling the height and weight of each drop to be the same so that each drop generates the same energy; and using positioning marks to ensure that the distance between the excitation point and the first detector in each measurement is constant so that the small offset distance is the same; Dividing the seismic signal into different time windows, each time window corresponding to a different depth range underground, including: obtaining a time-depth conversion relationship based on the average propagation speed of the seismic signal in the underground medium; dividing the seismic signal into multiple time windows according to preset time intervals; Based on the time-depth conversion relationship, a mapping relationship between each time window and the corresponding depth range is established. The mapping relationship is used for hierarchical identification of underground structures and disease types; wherein, stratification refers to underground structure layers of different depth ranges, each depth layer contains multiple signals, and each signal corresponds to data collected by detectors at different positions.

2. The underground disease detection method based on energy analysis according to claim 1, characterized in that: Calculate the energy spectrum of each signal in each time window, including: Preprocessing the seismic signals in each time window; According to the pre-processed seismic signal, the energy spectrum of each signal in each time window is calculated; The test section is divided according to the preset intervals, and the average energy value of each divided section is calculated.

3. The underground disease detection method based on energy analysis according to claim 2, characterized in that: Calculate the energy change rate of each signal, including: Select a known dense layer as a reference benchmark; Calculate the energy change rate P of each signal relative to the reference standard based on the average energy value of each divided segment; The underground structure density state of each divided section is divided according to the energy change rate.

4. The underground disease detection method based on energy analysis according to claim 3 is characterized in that: Calculate the energy change rate P of each signal relative to the reference standard using the following formula: P = (Em - Er) / Er Where P is the energy change rate, Em is the average energy value of the channel signal to be measured in the corresponding divided segment, and Er is the average energy value of the corresponding reference benchmark dense channel signal in the same divided segment.

5. The underground disease detection method based on energy analysis according to claim 3 is characterized in that: The underground structure density state of each section is divided according to the energy change rate, including: When the energy change rate is less than the threshold ε1, it is judged to be moderate to severe loose disease, and the density state is low density; When the energy change rate is greater than or equal to the threshold ε1 and less than the threshold ε2, it is determined to have abnormal characteristics of loose disease and the density state is medium density; When the energy change rate is greater than or equal to the threshold ε2 and less than the threshold ε3, it is determined to be a normal road section and the compaction state is compact; When the energy change rate is greater than or equal to the threshold ε3 and less than the threshold ε4, it is judged to be a small cavity or loose disease, and the density state is medium density; When the energy change rate is greater than the threshold ε4, it is judged as a large cavity or serious loose disease, and the density state is low.

6. The underground disease detection method based on energy analysis according to claim 5, characterized in that: The value range of ε1 is -50% to -40%; The value range of ε2 is -40% to -20%; The value range of ε3 is +10% to +20%; The value range of ε4 is +20% to +40%.

7. The underground disease detection method based on energy analysis according to claim 3, characterized in that: Perform Hilbert-Huang transform on the acquired seismic signal, including: Perform EMD decomposition on the acquired seismic signal to obtain the IMF component; Calculate the energy structure distribution diagram of the IMF component of each data channel. The energy structure distribution diagram reflects the energy proportion of IMF components of different frequencies. Perform Hilbert-Huang transform on the seismic signal to obtain the Hilbert marginal spectrum; The following characteristic parameters are extracted from the Hilbert marginal spectrum: (1) Maximum energy; (2) The instantaneous frequency corresponding to the maximum energy; (3) The decay rate of energy after reaching the maximum value; The extracted characteristic parameters are used as Hilbert marginal spectrum features to determine the disease type and location.

8. The underground disease detection method based on energy analysis according to claim 7, characterized in that: Based on the energy change rate P and Hilbert marginal spectrum characteristics, the type and location of underground diseases can be determined, including: When the energy change rate P is greater than or equal to the threshold Δ1, the corresponding disease type is judged to be a cavity; When the energy change rate P is less than the threshold Δ2, and the Hilbert marginal spectrum characteristics and the IMF component energy structure distribution characteristics meet any of the following conditions, the disease type is determined to be loose body: (1) The decrease in the maximum energy in the Hilbert marginal spectrum is greater than the threshold β1; (2) The instantaneous frequency increase corresponding to the maximum energy in the Hilbert marginal spectrum is greater than the threshold β2; (3) The decay rate of the energy in the Hilbert marginal spectrum after reaching its maximum value decreases by more than the threshold β3; (4) The similarity of the IMF component energy structure is less than the threshold β4; According to the mapping relationship between each time window and the corresponding depth range in step S2 and the division result of the underground structure density state in step S4, the location and depth range of the underground disease are determined.

9. The underground disease detection method based on energy analysis according to claim 8, characterized in that: The value range of β1 is 20% to 40%; The value range of β2 is 15% to 30%; The value range of β3 is 25% to 45%; The value range of β4 is: 35% to 45%.

10. An underground disease detection system based on energy analysis, characterized in that: include: a signal acquisition module for collecting seismic reflection wave signals through a geophone system, wherein the geophone system comprises a geophone array formed by a plurality of geophones arranged at predetermined equal intervals; The time window module divides the seismic signal into different time windows and establishes a mapping relationship between each time window and the corresponding depth range based on the time-depth conversion relationship; The energy spectrum module pre-processes the seismic signals in each time window, calculates the energy spectrum of each signal in each time window, and calculates the average energy value of each divided segment; The energy change rate module calculates the energy change rate P of each signal by comparing the energy spectrum of each signal in each time window with the energy spectrum of each signal in the known dense layer; Hilbert-Huang module, which performs empirical mode decomposition and Hilbert-Huang transform on the acquired seismic signals to obtain the IMF component energy structure distribution and Hilbert marginal spectrum characteristics; The disease judgment module judges the type and location of underground diseases based on the energy change rate P, the IMF component energy structure distribution and the Hilbert marginal spectrum characteristics, and determines the depth range of underground diseases based on the mapping relationship between the time window and the depth range.

Images