Drainage pipe clogging detection method and device based on ground penetrating radar, electronic equipment and storage medium

Abstract

The present application relates to the field of physical detection technology, and provides a sewer clogging detection method and device based on ground penetrating radar, an electronic device and a storage medium, the method comprising: constructing a sewer physical model under different clogging conditions, performing numerical simulation on the sewer physical model to obtain simulated radar maps corresponding to different clogging conditions; obtaining original radar data collected by the ground penetrating radar at a target sewer; removing direct waves from the original radar data and removing scattering noise in the original radar data according to an amplitude variance threshold to obtain actual radar data after noise removal, and generating an actual radar map based on the actual radar data; and matching and comparing the actual radar map with the simulated radar maps to determine a clogging detection result of the target sewer. The present application greatly improves the identification and interpretation stability of the radar map for small internal clogging in the sewer, and avoids the risk of damaging the tunnel bottom anti-seepage structure by conventional invasive means.

Description

Technical Field

[0001] This invention relates to the field of physical detection technology, and in particular to a method, apparatus, electronic device, and storage medium for detecting blockages in drainage pipes based on ground-penetrating radar. Background Technology

[0002] Drainage pipes are prone to internal blockage during long-term operation. Accurate detection of blockage in drainage pipes is of great significance to ensuring the safe operation of the pipes.

[0003] Existing technologies typically employ ground-penetrating radar (GPR) to emit electromagnetic waves and receive reflected signals for detection. However, this method is susceptible to interference from dense external reinforcement and covering layers, resulting in limited detection accuracy. Furthermore, it has poor ability to distinguish between water and silt within the pipe, making it difficult to accurately detect the specific degree of siltation. In addition, the detection results of this method heavily rely on manual interpretation, lack automated analysis methods, and have low efficiency in large-scale data processing. Summary of the Invention

[0004] This invention provides a method, device, electronic equipment, and storage medium for detecting blockages in drainage pipes based on ground-penetrating radar, in order to overcome the deficiencies existing in the prior art.

[0005] This invention provides a method for detecting blockages in drainage pipes based on ground-penetrating radar, comprising the following steps: Physical models of drainage pipes under different clogging conditions are constructed, and numerical simulations are performed on the physical models of drainage pipes to obtain simulated radar maps corresponding to the different clogging conditions. Acquire raw radar data collected by ground-penetrating radar at the target's drainage pipe; The direct wave is removed from the original radar data, and the scattering noise in the original radar data is removed according to the amplitude variance threshold to obtain the denoised actual radar data. The actual radar map is generated based on the actual radar data. The actual radar image is matched and compared with the simulated radar image to determine the blockage detection result of the target drainage pipe.

[0006] According to the present invention, a method for detecting blockages in drainage pipes based on ground-penetrating radar includes removing direct waves from the original radar data and removing scattering noise from the original radar data according to an amplitude variance threshold to obtain denoised actual radar data, comprising: From the raw radar data of multiple channels, the spatial and temporal distribution patterns of the direct wave are estimated by means of minimum error fitting, and the direct wave is subtracted from the raw radar data by combining a smooth window sequence to obtain preliminary denoised data; The abnormal amplitude intensity of the preliminary denoised data in each channel signal is statistically analyzed. Channel signals with a combined judgment threshold are regarded as scattering noise generated by local anomalies and the scattering noise is removed to obtain the actual radar data.

[0007] According to the present invention, a method for detecting blockages in drainage pipes based on ground-penetrating radar includes subtracting the direct wave from the original radar data using a smoothing window sequence to obtain preliminary denoised data, comprising: The arrival time of the direct wave and the signal duration of the direct wave are extracted from the channel signal where the original radar data is located. Based on the arrival time and the signal duration, a discrete time series with a cosine function as the envelope feature is constructed as the smoothing window sequence; The original radar data is weighted and subtracted from the smooth window sequence to eliminate the shallow direct wave energy in the original radar data, and the preliminary denoised data is output.

[0008] According to the present invention, a method for detecting blockages in drainage pipes based on ground-penetrating radar includes the following steps: statistically analyzing the abnormal amplitude intensity of the preliminary denoised data in each channel signal, and classifying channel signals exceeding a combined judgment threshold as scattering noise generated by local anomalies. Calculate the average amplitude of the preliminary denoised data across all channels, and calculate the variance of the amplitude distribution of the preliminary denoised data across all channels; The average amplitude is summed with the variance at a set multiple to obtain a benchmark value, which is then used as the combination judgment threshold. The abnormal amplitude intensity in each channel is compared with the combined judgment threshold. If the abnormal amplitude intensity in any channel is greater than the combined judgment threshold, the corresponding channel signal is extracted as the scattering noise.

[0009] According to the present invention, a method for detecting blockages in drainage pipes based on ground-penetrating radar includes matching and comparing the actual radar image with the simulated radar image to determine the blockage detection result of the target drainage pipe, comprising: If the matching result shows that there is a reflection feature with a complete hyperbola shape and no amplitude attenuation in the actual radar spectrum, then the blockage detection result is determined to be a non-blockage state in which the pipe is filled with air medium. If the matching result shows that there is a strong amplitude resonance signal at the deep time profile in the actual radar spectrum due to the reciprocating reflection and interference of electromagnetic waves in the water, then the siltation detection result is determined to be a water-filled state. If the matching result shows that the phase axis of the reflected wave in the actual radar spectrum is discontinuous and the energy of the reflected wave is attenuated compared to the unblocked state, then the blockage detection result is determined to be that there is silt or concrete blockage, and the blockage ratio is evaluated based on the degree of attenuation of the reflected wave energy.

[0010] According to the present invention, a method for detecting blockages in drainage pipes based on ground-penetrating radar includes assessing the blockage ratio based on the attenuation of reflected wave energy, comprising: Extract the attenuation degree of reflected wave energy in the actual radar spectrum, and extract the continuity characteristics of the phase axis of the reflected wave and the length of the horizontal distribution interval in the actual radar spectrum; The attenuation level, the continuity feature, and the length of the horizontal distribution interval are compared with the pre-established feature mapping relationship in the simulated radar spectrum to obtain the feature comparison result. The siltation ratio of the target drainage pipe is determined based on the feature comparison results; Specifically, when the feature comparison result indicates that the attenuation degree increases, the continuity feature deteriorates, and the length of the horizontal distribution interval shortens, the corresponding sludge ratio is determined to increase; when the feature comparison result indicates that the reflected wave energy attenuates to the point that the hyperbolic reflection feature at the top of the pipe disappears and only clutter remains, the corresponding sludge ratio is determined to be 100%.

[0011] According to the present invention, a method for detecting blockages in drainage pipes based on ground-penetrating radar includes, in which a physical model of the drainage pipe is numerically simulated to obtain simulated radar maps corresponding to different blockage conditions, comprising: Acquire the medium parameters of the target detection environment and the preset transmission center frequency of the ground-penetrating radar antenna; Calculate the wavelength of the electromagnetic wave in the detection medium based on the medium parameters and the transmission center frequency; If the set pipe wall thickness in the physical model of the drainage pipe is less than one-quarter of the wavelength, then the outer interface and the inner interface of the pipe wall are equivalent to a single composite reflection interface that cannot be independently layered, so as to simplify the pipe wall structure in the physical model of the drainage pipe. Based on the difference in relative permittivity between the different media on both sides of the single composite reflective interface, the interface reflection coefficient is derived, the propagation response of electromagnetic waves in the simplified physical model of the drainage pipe is solved, and the corresponding simulated radar spectrum is generated.

[0012] The present invention also provides a drainage pipe blockage detection device based on ground penetrating radar, comprising the following modules: The module is used to build physical models of drainage pipes under different clogging conditions, perform numerical simulations on the physical models of drainage pipes, and obtain simulated radar maps corresponding to the different clogging conditions. The acquisition module is used to acquire raw radar data collected by the ground-penetrating radar at the target drainage pipe; The denoising module is used to remove direct waves from the original radar data and remove scattering noise from the original radar data according to the amplitude variance threshold to obtain the denoised actual radar data, and generate an actual radar map based on the actual radar data. The detection module is used to match and compare the actual radar image with the simulated radar image to determine the blockage detection result of the target drainage pipe.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the drainage pipe blockage detection method based on ground penetrating radar as described above.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the drainage pipe blockage detection method based on ground penetrating radar as described above.

[0015] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the drainage pipe blockage detection method based on ground penetrating radar as described above.

[0016] This invention provides a ground-penetrating radar-based method, device, electronic equipment, and storage medium for detecting blockages in drainage pipes. By constructing a physical model of the drainage pipe encompassing various typical operating conditions and implementing finite-difference time-domain numerical simulation, combined with direct wave suppression and variance threshold denoising of actual radar data, it achieves interactive verification and matching comparison between high signal-to-noise ratio actual radar images and benchmark simulated images. Because the thin-walled structure is reasonably equivalently represented by dielectric parameter differences in the early stages, and strong point scattering interference from complex operating conditions is effectively suppressed during the detection and processing flow, the identification accuracy and interpretation stability of radar images for blockages inside small drainage pipes are significantly improved. This avoids the risk of damaging the tunnel's bottom seepage prevention structure using conventional invasive methods. Compared to similar non-destructive testing technologies, the accuracy of drainage pipe blockage detection is increased by more than 30%, and the error rate of blockage identification is reduced by 20%, providing a highly reliable automated interpretation basis for trenchless pipeline operation and maintenance. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the drainage pipe blockage detection method based on ground-penetrating radar provided by the present invention.

[0019] Figure 2 This is a schematic diagram of the cross-sectional structure of 11 types of drainage pipe blockages and blockage degree models provided by the present invention.

[0020] Figure 3 This is a schematic diagram of the drainage pipe blockage detection device based on ground-penetrating radar provided by the present invention.

[0021] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0023] The health condition of drainage pipes directly affects the safe and stable operation of the engineering seepage prevention and drainage system. For detecting blockages in drainage pipes beneath the concrete of water conveyance tunnels, common methods are mainly divided into two types: invasive pipe testing and non-invasive non-destructive testing.

[0024] Invasive pipeline inspection primarily relies on visual methods, including closed-circuit television (CCTV) pipeline inspection robots, pipeline periscopes, pipeline endoscopes, and underwater sonar detection. CCTV pipeline inspection robots use tracked robots equipped with high-definition cameras to enter the pipeline and transmit real-time images. Pipeline periscopes use retractable poles with high-definition cameras to quickly probe from wellheads or pipe openings. Sonar detection technology utilizes the propagation and reflection characteristics of ultrasonic waves in water to generate sediment topographic maps. On the other hand, non-invasive, non-destructive testing mainly includes geophysical methods such as ground-penetrating radar (GPR), electromagnetic induction, seismic wave methods, and resistivity imaging. GPR, a widely used method, locates and detects pipelines by emitting high-frequency electromagnetic waves and analyzing the reflected signals at the medium interface.

[0025] Despite the existence of various detection methods, these methods exhibit significant shortcomings when faced with the characteristics and special application scenarios of drainage pipes in water conveyance tunnels, such as small diameter, shallow burial depth, and double-layer reinforcement. Firstly, invasive pipe inspection methods are prone to failure. Due to the high requirements for seepage prevention in water conveyance tunnels, conventional pipe inspection often requires destructive methods to cut open the pipe, damaging the seepage prevention structure and drainage system of the tunnel floor. Simultaneously, the internal conditions of the pipe are complex, containing various blockages such as water, concrete blocks, and silt, significantly limiting the detection range and efficiency of visual inspection methods such as closed-circuit television pipeline inspection robots and pipeline periscopes. Secondly, for non-invasive non-destructive testing methods such as ground-penetrating radar and electromagnetic induction, the lack of effective dielectric parameter differences and the interference from double-layer reinforcement significantly affect the detection of drainage pipes, and also have considerable limitations in detecting the distribution of water and the degree of concrete blockage within the drainage pipe. In addition, the above methods generally have some common shortcomings: data processing relies on manual interpretation and lacks automated analysis systems, resulting in low processing efficiency and difficulty in dealing with large-scale detection data; professional equipment is expensive and requires professional personnel to operate, resulting in high detection costs and difficulty in large-scale popularization; a single technology is difficult to comprehensively assess the structural and functional defects of pipelines at the same time, and the concrete cover layer will attenuate the signal, further affecting the detection accuracy and depth.

[0026] To address this issue, this invention provides a ground-penetrating radar (GPR)-based method for detecting blockages in drainage pipes. The method involves constructing physical models of drainage pipes under different blockage conditions, performing numerical simulations of these models to obtain simulated radar maps corresponding to different blockage conditions, acquiring raw radar data collected by the GPR at the target drainage pipe, removing direct waves from the raw radar data, and removing scattering noise based on an amplitude variance threshold to obtain denoised actual radar data, generating an actual radar map based on the actual radar data, and matching and comparing the actual radar map with the simulated radar map to determine the blockage detection result of the target drainage pipe. This method achieves non-destructive, accurate, rapid, and automated identification of blockage conditions in drainage pipes under complex conditions such as water conveyance tunnels, effectively overcoming the shortcomings of traditional detection methods, such as damaging anti-seepage structures, being subject to interference from complex environments, and the low efficiency of manual interpretation. It meets the urgent engineering needs for large-scale pipeline network surveys and inspections.

[0027] in, Figure 1 This is a flowchart illustrating the drainage pipe blockage detection method based on ground-penetrating radar provided by the present invention, as shown below. Figure 1 As shown, the method includes steps 110, 120, 130 and 140.

[0028] Step 110: Construct physical models of drainage pipes under different clogging conditions, perform numerical simulations on the physical models of drainage pipes, and obtain simulated radar maps corresponding to different clogging conditions.

[0029] Here, different clogging conditions can be understood as the filling state and type of filler material inside the drainage pipe. These can include an air-filled state without clogging, a water-filled state, a semi-water state with a mixture of water and air, a pipe collapse and deformation state, and clogging states with different proportions of silt or concrete. For example, in the actual situation under large water conveyance tunnels, different clogging conditions can typically be an empty pipe state with 100% air inside; a semi-clogging state containing 50% solid concrete and 50% air; or a complete clogging state with the entire cross-section filled with solid concrete. The drainage pipe physical model refers to a medium structure model that reflects the real electromagnetic propagation characteristics, established in a virtual simulation space based on the geometric dimensions and material parameters of the actual engineering pipeline.

[0030] As an optional embodiment, the physical model of the drainage pipe can be constructed by obtaining the medium parameters of the target detection environment. For example, the relative permittivity, conductivity, and relative magnetic permeability of each medium can be pre-defined in the model. Specifically, the relative permittivity of air is 1 and its conductivity is 0; the relative permittivity of water is 80 and its conductivity is 0.08; the relative permittivity of the surrounding concrete is 8 and its conductivity is 0.01; and the relative permittivity of the PVC drainage pipe is 3 and its conductivity is 10. -6 The relative permittivity of the moist mud and sand is 12, and its conductivity is 0.5; the relative permittivity of the internal concrete, which acts as a blockage, is 10, and its conductivity is 0.02. Because these rigorous medium parameters are obtained, a physical model of the drainage pipe that accurately reflects the actual engineering conditions can be constructed based on these physical parameters.

[0031] Based on the physical model of the drainage pipe, and considering the thin wall of the actual PVC drainage pipe, this embodiment preferably adopts a thin-layer equivalent simplification method to avoid redundant model meshing and reduced computational efficiency caused by the thin-layer structure. Specifically, the transmission center frequency of the preset ground-penetrating radar antenna, such as 800MHz, is obtained, and then the wavelength of the electromagnetic wave in the concrete medium is calculated, which is approximately 13cm. Since the actual wall thickness of the PVC drainage pipe pre-embedded in the water conveyance tunnel is less than 1cm, which is much smaller than 1 / 4 wavelength, i.e., 3.8cm, under these physical conditions, the reflected signals from the top and bottom interfaces of the pipe wall will superimpose. The thin wall only exhibits a composite reflection interface, which does not have the physical basis for independent layered modeling. Based on this, in order to simplify the model, this embodiment treats the outer and inner interfaces of the pipe wall as equivalent to a single composite reflection interface that cannot be independently layered, thus simplifying the pipe wall structure in the physical model of the drainage pipe.

[0032] Furthermore, the interface reflection coefficient can be derived based on the difference in relative permittivity between the different media on both sides of a single composite reflective interface. Let the wave impedance be equal to the square root of the ratio of permeability to permittivity; the interface reflection coefficient is equal to the difference between the wave impedance of the second medium and the wave impedance of the first medium divided by their sum; the reflectivity is equal to the square of the absolute value of the reflection coefficient. Since the permittivity of PVC differs greatly from that of air and water, calculations show that the reflectivity of the air interface inside the pipe is 7.2% and that of the water interface is 45.6% in the unblocked state, forming a strong reflective interface with the pipe wall. However, the difference in permittivity between the concrete blockage and the concrete outside the pipe wall is smaller, resulting in a significant reduction in reflected wave energy.

[0033] Because a highly equivalent physical model of the drainage pipe has been constructed, Maxwell's equations can be solved using the finite-difference time-domain algorithm, enabling numerical simulation of the drainage pipe physical model. For example, ground-penetrating radar forward modeling software based on the finite-difference time-domain algorithm can be used to simulate the propagation process of electromagnetic waves in complex media. By setting 11 different typical clogging conditions for forward modeling, simulated radar maps corresponding to these 11 conditions are output, thereby establishing a standardized map feature response benchmark library.

[0034] Step 120: Obtain the raw radar data collected by the ground penetrating radar at the target drainage pipe.

[0035] Here, the target drainage pipe refers to the underlying water supply pipe in the actual engineering site that is to be tested for its health status. The raw radar data can be understood as the sequence of electromagnetic reflection echo signals collected by the receiving antenna after the ground penetrating radar emits high-frequency electromagnetic waves towards the target drainage pipe and the surrounding strata, without any filtering processing.

[0036] As an alternative embodiment, raw radar data can be acquired using a 1:1 scale simulation physical model. For example, a physical model can be created based strictly on a 250mm nominal diameter PVC drainage pipe used in actual numerical simulations, and sealing devices can be used at both ends of the model to simulate the closed environment under different filling conditions.

[0037] Because the quality of raw radar data is directly affected by the scanning equipment's moving speed and sampling frequency, operations can be performed based on pre-set high-density acquisition parameters to obtain highly continuous raw radar data. For example, using a ground-penetrating radar equipped with an 800MHz high-frequency shielded antenna, the radar antenna is controlled to move at a uniform speed of 0.1m / s along the axis of the target drainage pipe, with the sampling interval precisely set to 0.005m. Through this continuous profile scanning method, high-precision raw radar data is ultimately obtained.

[0038] Step 130: Remove the direct wave from the original radar data and remove the scattering noise from the original radar data according to the amplitude variance threshold to obtain the denoised actual radar data, and generate the actual radar map based on the actual radar data.

[0039] Here, direct wave can be understood as a high-energy electromagnetic wave that propagates directly from the transmitting transducer or along the shallow surface of the medium to the receiving transducer. Excessive energy can mask the effective reflection signal at early depths. Scattering noise refers to random, strong-point interference waves generated by obvious local physical anomalies on the concrete lining surface. Actual radar imagery refers to the conversion of denoised temporal and spatial sequence data into a two-dimensional or three-dimensional echo image for visual comparison.

[0040] Considering that generating a map based solely on raw radar data may contain a large amount of environmental clutter, causing weak reflection signals from blockage interfaces inside the pipeline to be obscured. For example, the strong amplitude of direct waves can mask target reflection waves from the top of shallow pipelines. Relying solely on the hardware filtering of the equipment itself may not be able to filter out localized scattering clutter generated by double-layered steel bars or aggregates within the medium, leading to distortion of subsequent radar waveforms. Therefore, this embodiment, in order to highlight the true target signals inside the pipeline, combines direct wave removal and variance threshold denoising, which can significantly improve the signal-to-noise ratio of the data and avoid misinterpretation.

[0041] As an optional implementation, the spatial and temporal distribution patterns of the direct wave can be estimated using an adaptive least squares fitting method based on the multi-channel signals of the original radar data. Subsequently, the arrival time and duration of the direct wave are extracted from the original channel signals, and a discrete smoothing sequence of the Hanning window function is constructed using the cosine function. Subtracting this discrete smoothing sequence from the original radar data achieves the removal of the direct wave.

[0042] As an optional implementation, after removing the direct wave, the abnormal amplitude intensity exhibited by the current data in each channel signal can be statistically analyzed. Simultaneously, the average amplitude and variance of the amplitude distribution across all channels are calculated. A combined judgment threshold is set by adding three times the variance to the average amplitude, and the abnormal amplitude intensity of each channel is compared sequentially. If the abnormal amplitude intensity of a certain channel exceeds the combined judgment threshold, the signal of that channel is extracted as scattering noise and removed, thus obtaining the denoised actual radar data.

[0043] Based on the established high signal-to-noise ratio actual radar data, further visualization and image enhancement processing are performed on the actual radar data to generate actual radar maps that reflect the characteristics of stratigraphic echoes.

[0044] Step 140: Match and compare the actual radar image with the simulated radar image to determine the detection results of the blockage in the target drainage pipe.

[0045] Here, matching and comparison can be understood as extracting key waveform features from the actual radar imagery and comparing them with benchmark features in a pre-established simulated radar imagery library using morphological and energetic similarity. Key waveform features can include the amplitude attenuation law of reflected waves, hyperbolic morphological changes, and differences in phase axis characteristics. The siltation detection result refers to the current internal state of the target drainage pipe, such as whether it contains silt, whether it is filled with water, whether there is pipe wall collapse or deformation, and the specific degree of siltation.

[0046] As an optional implementation, the presence of water in the pipe can be determined based on the amplitude and depth characteristics in the actual radar image. For example, if a strong amplitude resonance signal is detected at a deep location in the time profile of the actual radar image, and the reflected signal from the upper pipe wall exhibits multiple wave superposition, then analysis using the electromagnetic wave velocity formula reveals that this phenomenon is due to the fact that the propagation speed of electromagnetic waves in water is only 1 / 9 that in air. This significantly increases the time required for the wave beam to reach the bottom of the pipe and reflect back to the top, and also causes reciprocating reflection interference at the water boundary. Therefore, it can be determined that the target drainage pipe is filled with water.

[0047] As an optional embodiment, the blockage status of solid materials inside the pipe can be determined based on the continuity characteristics of the phase axis in the actual radar image. For example, if a reflection feature with a complete hyperbolic shape and no amplitude attenuation is found in the actual radar image, the blockage detection result of the target drainage pipe is determined to be a non-blockage state with the pipe filled with air. If the phase axis of the reflected wave at the top of the pipe in the actual radar image shows discontinuity, downward bending, or other characteristics, and the electromagnetic wave energy is attenuated due to the strong conductivity of the blockage, resulting in a decrease in the reflected wave energy compared to the non-blockage state, then it is determined that there is silt or concrete blockage, and the specific blockage ratio can be assessed based on the degree of attenuation of the reflected wave energy and the shortening of the horizontal distribution range.

[0048] Furthermore, considering that as the proportion of blockage to the pipe's cross-sectional area continues to increase, the difference in wave impedance between the blockage and the surrounding concrete becomes extremely small. To avoid misjudgment, this embodiment preferably adopts the limit feature determination method. When the effective signal reduction of the hyperbolic reflection interface located at the top of the pipe in the actual radar image is most significant, the in-phase axis of the reflection at the top of the pipe disappears, and only weak clutter or no effective reflection information remains, it can be determined that the target drainage pipe is in a state of complete cross-section blockage.

[0049] The ground-penetrating radar (GPR)-based drainage pipe blockage detection method provided in this embodiment constructs a physical model of the drainage pipe encompassing various typical operating conditions and implements time-domain finite-difference numerical simulation. Combined with direct wave suppression and variance threshold denoising of the actual acquired radar data, it achieves interactive verification and matching comparison between the high signal-to-noise ratio actual radar image and the benchmark simulated image. Because the thin-walled structure is reasonably equivalently represented by considering dielectric parameter differences in the early stages, and strong point scattering interference caused by complex operating conditions is effectively suppressed during the detection and processing flow, the identification accuracy and interpretation stability of the radar image for blockages inside small drainage pipes are significantly improved. This avoids the risk of damaging the tunnel's bottom seepage prevention structure using conventional invasive methods. Compared with similar non-destructive testing technologies, the accuracy of drainage pipe blockage detection is increased by more than 30%, and the error rate of blockage identification is reduced by 20%, providing a highly reliable automated interpretation basis for trenchless pipeline operation and maintenance.

[0050] Based on the above embodiments, direct waves are removed from the original radar data, and scattering noise in the original radar data is removed according to the amplitude variance threshold to obtain the denoised actual radar data, including: From the raw radar data of multiple channels, the spatial and temporal distribution patterns of the direct wave are estimated by the minimum error fitting method, and the direct wave is subtracted from the raw radar data by combining the smooth window sequence to obtain the preliminary denoised data. The abnormal amplitude intensity of the preliminary denoised data in each channel signal is statistically analyzed. Channel signals with a combined judgment threshold are regarded as scattering noise generated by local anomalies and are removed to obtain the actual radar data.

[0051] Here, raw radar data from multiple channels refers to the unprocessed echo sequence containing two-dimensional temporal and spatial information, acquired by ground-penetrating radar through multiple receiving antennas or multiple moving samples. Minimum error fitting can be understood as an algorithmic model such as adaptive least squares fitting, used to find the mathematical distribution that best fits the direct wave pattern. Spatial and temporal distribution patterns refer to the energy attenuation and phase change patterns of the direct wave at different receiving channel locations and at different receiving times. A smoothing window sequence refers to a mathematical weighting sequence used to smoothly transition and truncate the signal in the time domain, such as a Hanning window function sequence. Preliminary denoising data refers to transitional radar signals that have only had direct wave interference removed but still contain other random scattered clutter.

[0052] As an optional implementation, the spatial and temporal distribution patterns of the direct wave can be estimated from the original multi-channel radar data based on an adaptive least-squares fitting method. Subsequently, the arrival time and signal duration of the direct wave are extracted. Based on this, a smoothing window sequence constructed based on cosine function features is used to subtract the direct wave from the original radar data, obtaining preliminary denoised data. For example, suppose... The position sensor transmits signals that are The signal received by the sensor at that location is The sampling frequency is Its discrete sequence is The arrival time of the direct wave is estimated from the channel signal. The duration of the direct wave signal is The formula for subtracting the direct wave from the original radar data by combining the smooth window sequence is as follows: ; ; In the above formula, This represents the initial denoised discrete sequence of data after removing the direct wave. This represents a discrete sequence of raw radar data. Indicates the sensor's emission position. Indicates the sensor's receiving location. Represents a continuous-time variable. Indicates the index number of the discrete sequence. Indicates the sampling frequency of the radar equipment. Indicates the arrival time of the direct wave. Indicates the duration of the direct wave signal. This indicates the total number of data points corresponding to the direct wave signal. Represents pi (π). This represents the cosine function. Using the above formula and a smoothing window sequence, the direct waves at early depths can be smoothly subtracted from the original radar data, thus outputting preliminary denoised data.

[0053] After obtaining the preliminary denoising data, considering that there may be local physical anomalies on the lining surface, resulting in strong point scattering noise remaining in the preliminary denoising data, in order to avoid these noises appearing as false pipe wall or blockage reflection characteristics in the subsequently generated map, this embodiment statistically analyzes the abnormal amplitude intensity of the preliminary denoising data in each channel signal, and regards the channel signal with a value greater than the combined judgment threshold as scattering noise generated by local anomalies, and removes the scattering noise to obtain the actual radar data.

[0054] By using a set combination of judgment thresholds to statistically analyze and eliminate abnormal amplitude intensities of signals from each channel, strong point scattering interference caused by random environmental factors can be automatically and accurately filtered out, thereby improving the signal-to-noise ratio and imaging clarity of the final radar image.

[0055] Here, abnormal amplitude intensity refers to the amplitude value of a channel signal deviating from the average level of the overall signal energy at a specific time point. The combined judgment threshold refers to a dynamic judgment boundary set based on the statistical characteristics of the overall radar data to distinguish normal target echoes from noise interference. Scattering noise generated by local anomalies refers to the strong random scattered echoes that occur when ground-penetrating radar electromagnetic waves encounter non-target structures (such as coarse aggregate, gaps, etc.) inside or on the surface of concrete. Actual radar data refers to the high-quality, clean signal set ultimately used to generate radar maps after direct wave suppression and point scattering noise filtering.

[0056] As an optional embodiment, the combined judgment threshold can be determined by calculating the overall statistical parameters of the preliminary denoised data, and noise removal can then be performed. Specifically, the amplitude values ​​exhibited by the preliminary denoised data in all channel signals are extracted and statistically analyzed, the average amplitude distribution in all channels is calculated, and the variance of the overall amplitude distribution is calculated. Subsequently, the average amplitude is added to three times the variance, and the sum is used as the combined judgment threshold. The abnormal amplitude intensity in each channel is then compared with the combined judgment threshold. If the abnormal amplitude intensity in one or more channels is detected to be greater than the combined judgment threshold, the corresponding channel signal is determined to be scattering noise generated by a local anomaly. Finally, the extracted scattering noise is directly filtered out, resulting in high-purity actual radar data.

[0057] Based on any of the above embodiments, by combining a smooth window sequence to subtract the direct wave from the original radar data, preliminary denoised data is obtained, including: The arrival time and signal duration of the direct wave are extracted from the channel signal where the original radar data is located. Based on arrival time and signal duration, a discrete time series with cosine function as envelope feature is constructed as a smooth window sequence; The original radar data is weighted and subtracted from the smoothed window sequence to eliminate the direct wave energy in the shallow layer of the original radar data, and the preliminary denoised data is output.

[0058] Here, the channel signal containing the original radar data refers to the electromagnetic wave sequence containing amplitude and time information recorded by a specific receiving channel of a ground-penetrating radar. The arrival time of the direct wave refers to the time point at which the electromagnetic wave propagates directly from the transmitter to the receiver and causes the channel signal to generate its first significant peak or trough. The signal duration of the direct wave refers to the time span from the onset of the direct wave to its energy decaying to the background level.

[0059] As an optional embodiment, time parameters can be extracted using waveform extremum tracking and energy threshold determination. For example, the first wave peak exceeding a set background amplitude threshold is found in the channel signal containing the original radar data, and its corresponding time is determined as the arrival time of the direct wave. Subsequently, the energy attenuation envelope of the direct wave packet is tracked, and when the envelope energy drops to 5% of the peak energy, the corresponding time difference is recorded and extracted as the signal duration of the direct wave.

[0060] After extracting the arrival time and duration of the direct wave, considering that directly truncating the radar signal would introduce high-frequency oscillation artifacts and lead to data distortion, this embodiment constructs a discrete-time series with a cosine function as the envelope feature based on the arrival time and signal duration, serving as a smoothing window sequence. The method of constructing a discrete-time series with a cosine function as the envelope feature enables a smooth transition of the filtering window at the boundaries, effectively suppressing the truncation effect and ensuring the continuity and authenticity of the radar signal after removing the direct wave frequency band.

[0061] Here, a discrete-time series with a cosine function as its envelope feature refers to a set of numerical sequences generated using the smooth fluctuations of the cosine curve as the basis for weight allocation, such as a Hanning window function sequence. A smoothing window sequence refers to a mathematical window used for time-domain operations with the original signal to smooth and suppress signal energy over a specific time period.

[0062] As an optional implementation, the time parameter can be converted into a discrete number of points using the sampling frequency of the radar equipment, and then substituted into the cosine window function formula. Specifically, the sampling frequency of the ground-penetrating radar is obtained, the arrival time of the direct wave is added to the signal duration of the direct wave, and then multiplied by the sampling frequency to calculate the total number of data points corresponding to the direct wave signal. Based on this total number of data points and the constant pi, a discrete time series with a cosine function as the envelope feature is constructed, and this is used as a smoothing window sequence.

[0063] After constructing the smooth window sequence, considering that a customized smoothing filtering tool for the energy distribution characteristics of direct waves is already available, the original radar data and the smooth window sequence are then weighted and subtracted to eliminate the direct wave energy located in the shallow layer in the original radar data, and the preliminary denoised data is output.

[0064] Here, the weighted subtraction calculation can be understood as the operation of numerically canceling out the original radar data according to the weight ratio set by the smoothing window sequence. Shallow direct wave energy refers to high-amplitude electromagnetic wave energy concentrated in the top region of the time profile, interfering with the detection of shallow targets. Preliminary denoising data refers to transitional radar data from which direct waves have been filtered out but without subsequent point scattering noise processing.

[0065] Based on any of the above embodiments, the abnormal amplitude intensity exhibited by the preliminary denoised data in each channel signal is statistically analyzed, and channel signals exceeding the combined judgment threshold are considered as scattering noise generated by local anomalies, including: Calculate the average amplitude of the preliminary denoised data across all channels, and calculate the variance of the amplitude distribution of the preliminary denoised data across all channels; The baseline value is obtained by summing the average amplitude with the variance at a set multiple, and the baseline value is used as the combination judgment threshold. The abnormal amplitude intensity in each channel is compared with the combined judgment threshold. If the abnormal amplitude intensity in any channel is greater than the combined judgment threshold, the corresponding channel signal is extracted as scattering noise.

[0066] Here, preliminary denoising data refers to radar data sequences that still retain echoes from strata and anomalies after direct wave suppression processing. All channels refer to the set of all receiving channels formed by the continuous scanning of the ground-penetrating radar along the survey line. The average amplitude refers to the arithmetic average level of the overall signal energy. The variance of the amplitude distribution refers to the degree of dispersion of the signal amplitude of each channel from the average value, used to characterize the severity of signal fluctuations.

[0067] As an optional embodiment, the amplitude average of the preliminary denoised data across all channels can be calculated by traversing the data sequences of all received channels, accumulating and averaging the absolute values ​​of the amplitudes at discrete time points. Subsequently, based on this amplitude average, the mean of the sum of squares of the differences between the amplitude and the average value for each channel is calculated, thereby calculating the variance of the amplitude distribution of the preliminary denoised data across all channels.

[0068] After calculating the average amplitude and variance of the preliminary denoised data in all channels, considering that strong point scattering noise usually manifests as a few significantly high-energy spikes that deviate from the normal distribution, a single average value cannot effectively filter them out. Therefore, in this embodiment, the average amplitude and the variance at a set multiple are summed to obtain a reference value, which is then used as the combined judgment threshold.

[0069] Here, the set factor refers to a weighting coefficient set based on experience or experimental data, used to amplify the range of influence of variance. The benchmark value refers to a dynamic value calculated by fusing the average value and the weighted variance. The combined judgment threshold refers to the final determined numerical limit used to distinguish normal target echoes from noise interference.

[0070] As an optional implementation, the set factor can be set to 3 based on the outlier detection principle. Specifically, the variance of 3 times is added to the average amplitude to obtain a benchmark value that can encompass most normal data fluctuation ranges. This benchmark value is then used as a combined judgment threshold for filtering strong point scattering clutter.

[0071] After obtaining the combined judgment threshold, considering that there is already a clear and adaptive noise filtering standard, the abnormal amplitude intensity in each channel can be compared with the combined judgment threshold in turn. If the abnormal amplitude intensity in any channel is greater than the combined judgment threshold, the corresponding channel signal is extracted as scattered noise.

[0072] Here, the anomalous amplitude intensity within each channel refers to the sudden high-energy characteristic value exhibited by the corresponding channel signal at a specific sampling time point. Scattering noise generated by local anomalies refers to the strong random echo that occurs when ground-penetrating radar electromagnetic waves encounter non-target structures within concrete.

[0073] As an optional implementation, the input preliminary denoising data matrix can be traversed, and the relationship between the abnormal amplitude intensity in each channel and the combined judgment threshold can be compared sequentially. If the abnormal amplitude intensity in any channel is detected to be greater than the combined judgment threshold, the corresponding channel signal is determined to be data generated by a local anomaly and extracted as scattering noise. After extraction, this scattering noise can be removed, and finally clean actual radar data is output.

[0074] Based on any of the above embodiments, the actual radar imagery is matched and compared with the simulated radar imagery to determine the blockage detection results of the target drainage pipe, including: If the matching result shows a reflection feature with a complete hyperbola shape and no amplitude attenuation in the actual radar spectrum, then the blockage detection result is determined to be a non-blockage state in which the pipe is filled with air medium. If the matching result shows a strong amplitude resonance signal at the deep time profile in the actual radar image, formed by the reciprocating reflection and interference of electromagnetic waves in the water, then the siltation detection result is determined to be a water-filled state. If the matching result shows that the phase axis of the reflected wave in the actual radar spectrum is discontinuous and the energy of the reflected wave is attenuated compared with the non-blocked state, then the blockage detection result is determined to be that there is silt or concrete blockage, and the blockage ratio is evaluated based on the degree of attenuation of the reflected wave energy.

[0075] Here, the matching result refers to the conclusion of the similarity in morphology and energy characteristics between the actual radar image and the pre-established simulated radar image. A complete hyperbolic morphology means that when the radar scans along the pipeline axially across the pipe top interface, the resulting continuous and smooth arched reflection waveform curve shows no damage or distortion. No amplitude attenuation means that the energy intensity of the reflected wave remains at a high initial reference level. A non-clogging state with the pipe filled with air refers to the ideal unobstructed condition where the pipe's cross-sectional space is 100% occupied by air.

[0076] As an optional embodiment, the reflection signal from the wall of the target drainage pipe in the actual radar image can be extracted. Based on relevant medium parameters, the relative permittivity of air is 1, while that of polyvinyl chloride (PVC) is 3. The significant difference in wave impedance between the two creates a strong reflection interface. If the actual radar image is compared with the simulated radar image corresponding to the empty pipe model, and the matching result shows a complete hyperbolic reflection characteristic in the actual radar image with no amplitude attenuation, it can be determined that the internal structure of the target drainage pipe has not changed, thus confirming that the blockage detection result indicates a non-blockage state with the pipe filled with air.

[0077] After ruling out the non-blocked state where the pipe is filled with air, considering that underground drainage pipes often accumulate water due to leakage or poor downstream drainage, the water body will significantly change the propagation speed and reflection path of electromagnetic waves. Therefore, if the matching result shows a strong amplitude resonance signal formed by the reciprocating reflection interference of electromagnetic waves in the water body at the deep time profile in the actual radar spectrum, then the blockage detection result is determined to be a water-filled state.

[0078] Here, the deep time profile refers to the display area with larger values ​​on the longitudinal time axis of the radar two-dimensional image, representing a longer two-way travel time of electromagnetic waves. Electromagnetic wave reciprocating reflection interference in water refers to the physical process where, after entering the water body inside the pipe, some of the energy bounces back and forth multiple times between the top and bottom interfaces of the pipe, and each wave undergoes constructive interference at the moment of reception. Strong amplitude resonant signal refers to the radar echo characteristic characterized by a significantly enhanced total amplitude due to the superposition of multiple wave interferences.

[0079] As an alternative embodiment, the analysis can be based on the fact that water has a relative permittivity of 80, a value that differs significantly from that of PVC material and the surrounding concrete. According to Fresnel's law of reflection, the greater the difference in permittivity, the stronger the reflection coefficient. Since the propagation speed of electromagnetic waves in water is approximately 1 / 9 that in air, the total time required for the electromagnetic wave to travel from the top of the pipe into the water, reach the bottom of the pipe, and then reflect back to the top is significantly increased, causing the echo signal to appear at a deeper part of the time profile. If the actual radar image is compared with the simulated radar image corresponding to the water-filled pipe model, and the matching result shows a strong amplitude resonance signal at a deeper time profile in the actual radar image due to the interference of electromagnetic waves reflecting back and forth in the water, then the blockage detection result is determined to be a water-filled state.

[0080] If the matching result shows that the phase axis of the reflected wave in the actual radar spectrum is discontinuous and the energy of the reflected wave is attenuated compared with the non-blocked state, then the blockage detection result is determined to be that there is silt or concrete blockage, and the blockage ratio is evaluated based on the degree of attenuation of the reflected wave energy.

[0081] Here, the reflected wave phase axis refers to the line connecting peaks or troughs with the same phase on adjacent channels in the radar spectrum. Discontinuity characteristics refer to the appearance of broken, misaligned, or distorted irregularities in what was originally a continuous hyperbola or smooth curve. Attenuation of reflected wave energy compared to the unblocked state means that the peak amplitude of the electromagnetic echo in the target area is significantly lower than the peak amplitude in the empty pipe reference state. Silt or concrete blockage refers to the pipe interior being filled with solid or semi-solid materials whose relative permittivity is close to that of the external structure. Blockage ratio refers to the percentage of the pipe's cross-sectional area occupied by the blockage material.

[0082] As an optional embodiment, parameters of the relevant solid medium are obtained. For example, the relative permittivity of silted concrete is 10 and its conductivity is 0.02, while that of moist mud and sand is 12 and its conductivity is 0.5. Since the permittivity of these silt deposits is relatively close to that of the concrete on the outer side of the pipe wall (8), the overall reflection coefficient is significantly lower. Furthermore, the silt and sand deposits after immersion exhibit high conductivity, which strongly absorbs and attenuates electromagnetic wave energy. If the actual radar spectrum is compared with the simulated radar spectrum, and the matching result shows that the phase axis of the reflected wave in the actual radar spectrum is discontinuous, and the reflected wave energy is attenuated compared to the unblocked state, then the siltation detection result is determined to be silt or concrete blockage. Simultaneously, the amplitude reduction and the shortening of the horizontal distribution range in the actual radar spectrum can be extracted and quantitatively mapped to the simulated radar spectra corresponding to 11 preset drainage pipe siltation degree models.

[0083] For example, features can be mapped to a semi-clogged model containing 50% concrete and 50% air, or a severely clogged model containing 75% concrete and 25% air, and the clogged proportion can be assessed based on the attenuation of reflected wave energy.

[0084] Based on any of the above embodiments, assessing the blockage ratio based on the attenuation of reflected wave energy includes: Extract the attenuation degree of reflected wave energy in the actual radar spectrum, and extract the continuity characteristics of the phase axis of the reflected wave and the length of the horizontal distribution interval in the actual radar spectrum; The attenuation level, continuity characteristics, and length of the horizontal distribution interval are compared with the pre-established feature mapping relationship in the simulated radar spectrum to obtain the feature comparison results. The siltation ratio of the target drainage pipe is determined based on the feature comparison results; Specifically, when the feature comparison results indicate an increase in attenuation, a deterioration in continuity, and a shortening of the horizontal distribution interval, the corresponding blockage ratio is determined to be increased; when the feature comparison results indicate that the reflected wave energy has attenuated to the point where the hyperbolic reflection characteristics at the top of the pipe disappear and only clutter remains, the corresponding blockage ratio is determined to be 100%.

[0085] Here, the pre-established feature mapping relationship refers to the mathematical correspondence rules between radar spectrum feature change parameters and the percentage of blockage in the pipe, summarized by performing radar forward modeling numerical simulations on physical models with different proportions such as 50% blockage, 75% blockage, 80% blockage, and 100% full-section blockage. The feature comparison result refers to the similarity interval or relative position coordinates matched by the actual extracted spectrum feature values ​​in the above mapping rules.

[0086] As an optional implementation, a database of radar imagery for 11 typical operating conditions, pre-calculated using a finite-difference time-domain algorithm, can be retrieved. The extracted attenuation level, continuity features, and horizontal distribution interval length are input into the pattern recognition module to calculate the cosine similarity between the current actual multidimensional feature combination and the feature combinations of each standard model in the database. The feature map of the target model with the highest similarity score is selected as the output, yielding the feature comparison result.

[0087] After obtaining the feature comparison results, considering that different degrees of blockage will cause progressive damage to the radar reflection interface, in order to intuitively output the final engineering diagnosis conclusion, the blockage ratio of the target drainage pipe is determined based on the feature comparison results. Specifically, when the feature comparison results indicate an increase in attenuation, a deterioration in continuity features, and a shortening of the horizontal distribution interval, the corresponding blockage ratio is determined to be increased. When the feature comparison results indicate that the reflected wave energy has attenuated to the point that the hyperbolic reflection feature at the top of the pipe disappears and only clutter remains, the corresponding blockage ratio is determined to be 100%.

[0088] Here, the blockage ratio refers to the percentage of the cross-sectional area of ​​the pipe occupied by solid blockage material to the total effective cross-sectional area of ​​the pipe. The hyperbolic reflection characteristic at the pipe top refers to the standard parabolic bright echo formed by the radar antenna across the top of a circular pipe when there is no blockage or only slight blockage. Residual clutter refers to the disordered and weak electromagnetic fluctuations received by the radar after the loss of effective medium interface reflection, caused by background environment and minor material inhomogeneities.

[0089] As an optional embodiment, the decision logic can be executed based on the physical evolution laws revealed in numerical simulations. Since the dielectric constant of the blockage (e.g., concrete, silt) is relatively small (e.g., the relative dielectric constant of concrete blockage is 10) and the relative dielectric constant of the concrete outside the pipe wall (value 8), this small difference in wave impedance causes a continuous deterioration in the continuity of the reflected signal's phase axis at the upper pipe wall, and a continuous shortening of the horizontal distribution range of the reflected signal, as the proportion of blockage in the pipe's cross-sectional area increases. Therefore, when the feature comparison results indicate an increase in attenuation, a deterioration in continuity, and a shortening of the horizontal distribution range, the corresponding increase in the blockage proportion can be determined (e.g., a transition from 75% blockage to 80% blockage). Furthermore, the blockage after immersion in water exhibits a high electrical conductivity due to its high water content, resulting in a strong absorption effect on electromagnetic wave energy. When the blockage fills the entire pipe space, the reflected wave amplitude of deep targets is significantly reduced or even submerged in background noise. Therefore, when the feature comparison results indicate that the reflected wave energy has decayed to the point where the hyperbolic reflection feature at the top of the pipe disappears and only clutter remains, i.e., when the radar profile shows a phenomenon with almost no reflection interface, the corresponding blockage ratio can be accurately determined to be 100%, thus providing a quantitative and intuitive basis for on-site pipeline dredging decisions.

[0090] Based on any of the above embodiments, a numerical simulation of the physical model of the drainage pipe is performed to obtain simulated radar maps corresponding to different clogging conditions, including: Acquire the medium parameters of the target detection environment and the preset transmission center frequency of the ground-penetrating radar antenna; Calculate the wavelength of the electromagnetic wave in the detection medium based on the medium parameters and the transmission center frequency; If the wall thickness in the physical model of the drainage pipe is less than one-quarter of the wavelength, the outer and inner interfaces of the pipe wall are equivalent to a single composite reflection interface that cannot be independently layered, so as to simplify the pipe wall structure in the physical model of the drainage pipe. Based on the difference in relative permittivity between the different media on both sides of a single composite reflective interface, the interface reflection coefficient is derived. The propagation response of electromagnetic waves in the simplified physical model of the drainage pipe is then solved, and the corresponding simulated radar spectrum is generated.

[0091] Here, the medium parameters of the target detection environment refer to the electromagnetic properties such as the relative permittivity, conductivity, and relative magnetic permeability of the concrete lining surrounding the target drainage pipe, the internal air, water, and various blockages. The preset transmission center frequency of the ground-penetrating radar antenna refers to the peak value of the main frequency of the high-frequency electromagnetic waves emitted by the detection equipment into the ground.

[0092] As an optional embodiment, the medium parameters of the target detection environment can be obtained through on-site sampling tests or by consulting standard physical parameter tables, and the hardware operating parameters of the detection equipment can be read. For example, the relative permittivity of the surrounding concrete is obtained as 8, the conductivity as 0.01, and the relative permeability as 1; the relative permittivity of the PVC drainage pipe is obtained as 3, the conductivity as 10⁻⁶, and the relative permeability as 1; and the relative permittivity of air is obtained as 1, the conductivity as 0, and the relative permeability as 1. Simultaneously, the transmission center frequency of the preset ground-penetrating radar antenna is obtained as 800MHz. Specifically, to ensure the accuracy and scientific validity of the radar forward modeling numerical simulation, the relevant medium parameters of the target detection environment obtained in this embodiment are shown in Table 1: Table 1 Relevant Media Parameters After obtaining the medium parameters of the target detection environment and the preset transmission center frequency of the ground-penetrating radar antenna, considering that the wavelength of electromagnetic waves directly determines the longitudinal resolution of the radar system, in order to determine whether the pipe wall structure will produce a layered reflection effect on the propagation of electromagnetic waves, this embodiment calculates the wavelength of electromagnetic waves in the detection medium based on the medium parameters and the transmission center frequency. Here, the wavelength of electromagnetic waves in the detection medium refers to the physical distance traversed by an electromagnetic wave of a specific frequency propagating for one complete cycle in a geological medium with a specific dielectric constant.

[0093] As an optional embodiment, the propagation speed of electromagnetic waves in the medium can be obtained by dividing the speed of light by the square root of the dielectric constant of the target medium. Then, by dividing the propagation speed of electromagnetic waves by the transmission center frequency, the wavelength of the electromagnetic wave in the detection medium can be calculated. For example, by substituting the relative dielectric constant of the surrounding concrete (8) and the transmission center frequency (800MHz) into the wavelength calculation formula, the wavelength of the electromagnetic wave in the detection medium is calculated to be 13cm.

[0094] After calculating that the wavelength of electromagnetic waves in the detection medium is 13cm, considering that the actual wall thickness of the PVC drainage pipes pre-buried in the water conveyance tunnel is less than 1cm, the set wall thickness is far less than the resolution limit of 1 / 4 wavelength. This results in the superposition of the reflected signals from the top and bottom interfaces of the pipe wall in the radar profile, which cannot be identified by the radar alone.

[0095] If the wall thickness in the physical model of the drainage pipe is less than 1 / 4 of the wavelength, the outer and inner interfaces of the pipe wall are equivalent to a single composite reflection interface that cannot be independently layered, thus simplifying the pipe wall structure in the physical model. Here, the wall thickness refers to the pre-configured radial physical thickness of the pipe shell in the physical model. A single composite reflection interface that cannot be independently layered refers to an equivalent boundary where, from the perspective of electromagnetic wave propagation, there is no physical basis for independent layered modeling, and the upper and lower physical interfaces are merged and regarded as a whole to produce a comprehensive reflection effect.

[0096] As an optional embodiment, the calculated wavelength of 13cm can be divided by 4 to obtain 1 / 4 of the wavelength, which is 3.8cm. If the system determines that the wall thickness in the physical model of the drainage pipe is less than 1 / 4 of the wavelength, i.e., 3.8cm, the electromagnetic field within the layer exists in the form of an evanescent wave, exhibiting a localized field whose amplitude decays exponentially with depth and does not propagate energy. Under this condition, the plane wave approximation can still estimate the reflection coefficient relatively well, providing a theoretical basis for the simplification of thin layers. Therefore, the system automatically treats the outer and inner interfaces of the pipe wall as equivalent to a single composite reflection interface that cannot be independently layered, thus simplifying the pipe wall structure in the physical model of the drainage pipe.

[0097] After simplifying the pipe wall structure in the physical model of the drainage pipe, considering that the difference in the comprehensive reflection coefficient between models including and excluding the pipe wall is small, and that the reflected waves generated on the surface of the pipe wall medium will superimpose with the main reflected waves, there is almost no difference in electromagnetic wave reflection characteristics. In order to accurately simulate the detection signal, the interface reflection coefficient is derived based on the difference in relative permittivity of the different media on both sides of the single composite reflection interface. The propagation response of electromagnetic waves in the simplified physical model of the drainage pipe is then solved to generate the corresponding simulated radar spectrum.

[0098] Here, the different media on both sides of a single composite reflective interface refer to the lining concrete medium on the outside of the equivalent interface and the air, water, or concrete blockage medium on the inside of the equivalent interface. The interface reflection coefficient is the ratio of the parameters of the reflected wave to the incident wave when an electromagnetic wave passes through the interface of two different media. The propagation response refers to the composite electromagnetic signal that finally reaches the receiving antenna after the electromagnetic wave undergoes transmission, reflection, attenuation, and interference in the medium.

[0099] As an optional embodiment, the interface reflection coefficient and reflectivity can be derived using the wave impedance formula. The specific calculation formula is as follows: ; ; ; In the above formula, Indicates the wave impedance of the medium. Indicates the magnetic permeability of the medium. Indicates the dielectric constant of the medium; Indicates the interface reflectance coefficient. This represents the wave impedance of the first medium on the incident side of the electromagnetic wave. This represents the wave impedance of the second medium on the electromagnetic wave transmission side; This represents the reflectivity of the interface.

[0100] Based on the above formula derivation, it can be seen that the dielectric constant of polyvinyl chloride (PVC) differs significantly from that of media such as air and water. In the unblocked state, with the pipe filled with either air or water, the reflectivity is calculated to be 7.2% when the pipe is filled with air and 45.6% when filled with water, forming a strong reflective interface with the pipe wall. The dielectric constant of underwater mud sludge sludge is less different from that of the outer concrete, resulting in a significant reduction in reflected wave energy. Subsequently, forward modeling software based on the two-dimensional finite-difference time-domain method was used to solve the propagation response of electromagnetic waves in the simplified physical model of the drainage pipe, ultimately generating simulated radar maps corresponding to 11 different scales and media of sludge sludge conditions.

[0101] After clarifying the derivation rules of various medium parameters and reflection coefficients, in order to ensure that the generated simulated radar map can fully cover various complex conditions that may occur inside the actual water conveyance tunnel, this embodiment specifically constructs 11 typical working condition models for the simplified drainage pipe physical model. The relevant parameters of the models are shown in Table 2: Table 2 Model-related parameters Based on the descriptions of various working conditions and the proportion of media inside the pipe set in Table 2, the propagation response of electromagnetic waves was solved for each of the 11 models using the finite-difference time-domain forward modeling software. Finally, a set of simulated radar maps covering different filling materials, different degrees of blockage, and pipe wall deformation conditions was generated.

[0102] After clarifying the relevant parameters of each model listed in Table 2, in order to more intuitively demonstrate the spatial distribution and cross-sectional shape of the medium inside the pipe in the above 11 typical working condition models, a detailed construction can be carried out in conjunction with the physical structure diagram. Specifically, Figure 2 This is a schematic cross-sectional view of the 11 types of drainage pipe blockages and their degree of blockage provided by this invention, as shown in the figure. Figure 2 As shown, these 11 physical models of drainage pipes cover various typical scenarios that may be encountered on-site in water conveyance tunnels: Among them, Model 1 to Model 3 represent the basic states of a circular pipe without structural deformation, corresponding to the empty pipe state with 100% air, the full water state with 100% water, and the semi-water state with water in the lower half and air in the upper half.

[0103] Models 4 and 5 demonstrate the geometric shape of the top collapse deformation of the drainage pipe under external load, and respectively correspond to the unblocked deformation state with air inside the pipe and the deformation state with water inside the pipe.

[0104] Models 6 to 9 visually demonstrate the gradual evolution of a circular pipe being blocked by a solid, concrete-like substance. These correspond to the following states: the bottom half of the pipe diameter is blocked by concrete (50% blockage); the top three-quarters of the pipe diameter is blocked by concrete (75% blockage); the pipe space is almost completely blocked by concrete (80% blockage); and the entire cross-section is completely blocked by concrete. In Models 6 to 8, the remaining space at the top that is not occupied by concrete is filled with air.

[0105] Models 10 and 11 further demonstrate a more complex mixed blockage condition accompanied by groundwater seepage. Model 10 corresponds to a semi-blocked and fully watered state in which the lower half of the pipe diameter is blocked by solid concrete and the remaining space in the upper half is filled with water. Model 11 corresponds to a complex blockage and fully watered state in which a layer of silt with an uneven surface is sandwiched between the lower solid concrete layer and the upper water body.

[0106] based on Figure 2 The 11 types of high-fidelity drainage pipe physical models shown are defined by their geometric boundaries, and each region is strictly assigned the relative permittivity, conductivity, and other attribute parameters set in Table 1 above. Maxwell's equations are then used to solve for the propagation response of electromagnetic waves in the simplified drainage pipe physical models, ultimately generating simulated radar maps covering different geometric shapes and complex filling media. This process provides a realistic and parameter-complete theoretical map library for accurate matching with subsequent actual radar data and quantitative assessment of complex blockage ratios.

[0107] The following describes the ground-penetrating radar-based drainage pipe blockage detection device provided by the present invention. The ground-penetrating radar-based drainage pipe blockage detection device described below and the ground-penetrating radar-based drainage pipe blockage detection method described above can be referred to in correspondence with each other.

[0108] Based on any of the above embodiments Figure 3 This is a schematic diagram of the drainage pipe blockage detection device based on ground penetrating radar provided by the present invention, as shown below. Figure 3 As shown, the device includes: Module 310 is used to build physical models of drainage pipes under different clogging conditions, perform numerical simulations on the physical models of drainage pipes, and obtain simulated radar maps corresponding to different clogging conditions. The acquisition module 320 is used to acquire the raw radar data collected by the ground penetrating radar at the target drainage pipe; The denoising module 330 is used to remove direct waves from the original radar data and remove scattering noise from the original radar data according to the amplitude variance threshold to obtain the denoised actual radar data, and generate an actual radar map based on the actual radar data. The detection module 340 is used to match and compare the actual radar spectrum with the simulated radar spectrum to determine the detection results of the blockage in the target drainage pipe.

[0109] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440. The processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions from the memory 430 to execute a ground-penetrating radar-based drainage pipe blockage detection method.

[0110] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0111] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the ground-penetrating radar-based drainage pipe blockage detection method provided by the above methods.

[0112] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the ground-penetrating radar-based drainage pipe blockage detection method provided by the above methods.

[0113] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0114] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting blockages in drainage pipes based on ground-penetrating radar, characterized in that, include: Physical models of drainage pipes under different clogging conditions are constructed, and numerical simulations are performed on the physical models of drainage pipes to obtain simulated radar maps corresponding to the different clogging conditions. Acquire raw radar data collected by ground-penetrating radar at the target's drainage pipe; The direct wave is removed from the original radar data, and the scattering noise in the original radar data is removed according to the amplitude variance threshold to obtain the denoised actual radar data. The actual radar map is generated based on the actual radar data. The actual radar image is matched and compared with the simulated radar image to determine the blockage detection result of the target drainage pipe.

2. The method for detecting blockages in drainage pipes based on ground-penetrating radar according to claim 1, characterized in that, The step of removing direct waves from the original radar data and removing scattering noise from the original radar data according to an amplitude variance threshold to obtain denoised actual radar data includes: From the raw radar data of multiple channels, the spatial and temporal distribution patterns of the direct wave are estimated by means of minimum error fitting, and the direct wave is subtracted from the raw radar data by combining a smooth window sequence to obtain preliminary denoised data; The abnormal amplitude intensity of the preliminary denoised data in each channel signal is statistically analyzed. Channel signals with a combined judgment threshold are regarded as scattering noise generated by local anomalies and the scattering noise is removed to obtain the actual radar data.

3. The drainage pipe blockage detection method based on ground-penetrating radar according to claim 2, characterized in that, The combined smoothing window sequence subtracts the direct wave from the original radar data to obtain preliminary denoised data, including: The arrival time of the direct wave and the signal duration of the direct wave are extracted from the channel signal where the original radar data is located. Based on the arrival time and the signal duration, a discrete time series with a cosine function as the envelope feature is constructed as the smooth window sequence; The original radar data is weighted and subtracted from the smooth window sequence to eliminate the direct wave energy in the shallow layer of the original radar data, and the preliminary denoised data is output.

4. The method for detecting blockages in drainage pipes based on ground-penetrating radar according to claim 2, characterized in that, The statistical analysis of the abnormal amplitude intensity exhibited by the preliminary denoised data in each channel signal, and the channel signal exceeding the combined judgment threshold as scattering noise generated by local anomalies, includes: Calculate the average amplitude of the preliminary denoised data across all channels, and calculate the variance of the amplitude distribution of the preliminary denoised data across all channels; The average amplitude is summed with the variance at a set multiple to obtain a benchmark value, which is then used as the combination judgment threshold. The abnormal amplitude intensity in each channel is compared with the combined judgment threshold. If the abnormal amplitude intensity in any channel is greater than the combined judgment threshold, the corresponding channel signal is extracted as the scattering noise.

5. The method for detecting blockages in drainage pipes based on ground-penetrating radar according to any one of claims 1 to 4, characterized in that, The step of matching and comparing the actual radar image with the simulated radar image to determine the blockage detection result of the target drainage pipe includes: If the matching result shows that there is a reflection feature with a complete hyperbola shape and no amplitude attenuation in the actual radar spectrum, then the blockage detection result is determined to be a non-blockage state in which the pipe is filled with air medium. If the matching result shows that there is a strong amplitude resonance signal at the deep time profile in the actual radar spectrum due to the reciprocating reflection and interference of electromagnetic waves in the water, then the siltation detection result is determined to be a water-filled state. If the matching result shows that the phase axis of the reflected wave in the actual radar spectrum is discontinuous and the energy of the reflected wave is attenuated compared to the unblocked state, then the blockage detection result is determined to be that there is silt or concrete blockage, and the blockage ratio is evaluated based on the degree of attenuation of the reflected wave energy.

6. The drainage pipe blockage detection method based on ground-penetrating radar according to claim 5, characterized in that, The assessment of the blockage ratio based on the attenuation of reflected wave energy includes: Extract the attenuation degree of reflected wave energy in the actual radar spectrum, and extract the continuity characteristics of the phase axis of the reflected wave and the length of the horizontal distribution interval in the actual radar spectrum; The attenuation level, the continuity feature, and the length of the horizontal distribution interval are compared with the pre-established feature mapping relationship in the simulated radar spectrum to obtain the feature comparison result. The siltation ratio of the target drainage pipe is determined based on the feature comparison results; Specifically, when the feature comparison result indicates that the attenuation degree increases, the continuity feature deteriorates, and the length of the horizontal distribution interval shortens, the corresponding sludge ratio is determined to increase; when the feature comparison result indicates that the reflected wave energy attenuates to the point that the hyperbolic reflection feature at the top of the pipe disappears and only clutter remains, the corresponding sludge ratio is determined to be 100%.

7. The method for detecting blockages in drainage pipes based on ground-penetrating radar according to any one of claims 1 to 4, characterized in that, The numerical simulation of the physical model of the drainage pipe to obtain simulated radar maps corresponding to different clogging conditions includes: Acquire the medium parameters of the target detection environment and the preset transmission center frequency of the ground-penetrating radar antenna; Calculate the wavelength of the electromagnetic wave in the detection medium based on the medium parameters and the transmission center frequency; If the set pipe wall thickness in the physical model of the drainage pipe is less than one-quarter of the wavelength, then the outer interface and the inner interface of the pipe wall are equivalent to a single composite reflection interface that cannot be independently layered, so as to simplify the pipe wall structure in the physical model of the drainage pipe. Based on the difference in relative permittivity between the different media on both sides of the single composite reflective interface, the interface reflection coefficient is derived, the propagation response of electromagnetic waves in the simplified physical model of the drainage pipe is solved, and the corresponding simulated radar spectrum is generated.

8. A drainage pipe blockage detection device based on ground-penetrating radar, characterized in that, include: The module is used to build physical models of drainage pipes under different clogging conditions, perform numerical simulations on the physical models of drainage pipes, and obtain simulated radar maps corresponding to the different clogging conditions. The acquisition module is used to acquire raw radar data collected by the ground penetrating radar at the target drainage pipe; The denoising module is used to remove direct waves from the original radar data and remove scattering noise from the original radar data according to the amplitude variance threshold to obtain the denoised actual radar data, and generate an actual radar map based on the actual radar data. The detection module is used to match and compare the actual radar image with the simulated radar image to determine the blockage detection result of the target drainage pipe.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the ground-penetrating radar-based drainage pipe blockage detection method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the ground-penetrating radar-based drainage pipe blockage detection method as described in any one of claims 1 to 7.

Images