Mixing point verification method and system based on sonar detection and positioning algorithm

Abstract

The application provides a method and system for verifying a mixed connection point based on a sonar detection and positioning algorithm, and relates to the technical field of sewer network detection. The method comprises the following steps: obtaining a first echo signal set of a target pipe section under a first working condition of fine weather without rain or known no mixed connection, extracting echo spectrum features to generate a reference echo spectrum feature model; obtaining a second echo signal set of the target pipe section under a second working condition of monitoring a triggering rain and sewage mixed connection event; comparing and analyzing echo spectrum features of the second echo signal set with the reference model to identify abnormal point positions of abnormal changes in the echo spectrum features; performing spatial clustering analysis on a plurality of abnormal point positions to generate abnormal point position clusters, determining a center position or a boundary position of each abnormal point position cluster as a candidate position of a mixed connection point, and generating a mixed connection point verification result after format processing. Through the comparison of sonar spectrum features and spatial clustering analysis, the application realizes accurate positioning of the mixed connection point and improves the detection efficiency and accuracy.

Description

Technical Field

[0001] This invention relates to the field of drainage pipe network detection technology, and in particular to a method and system for verifying mixed connection points based on sonar detection and positioning algorithms. Background Technology

[0002] In urban drainage network management, combined sewer overflows are a major cause of reduced sewage treatment efficiency and increased environmental pollution. Traditional methods for detecting overflow points typically rely on manual inspections, closed-circuit television (CCTV) monitoring, or flow monitoring. However, these methods have significant shortcomings: manual inspections are inefficient and struggle to cover complex networks; CCTV monitoring is limited by water transparency and the internal environment of the pipes, leading to misjudgments or missed detections; and flow monitoring only reflects overall flow changes and cannot accurately pinpoint the exact location of the overflow point. Furthermore, existing sonar detection technology is primarily used for pipe structure inspection and has not been effectively applied to the identification and verification of overflow points. Therefore, how to quickly and accurately locate overflow points under complex operating conditions has become a pressing technical problem in the current operation and maintenance management of drainage networks.

[0003] To address the aforementioned problems, this invention proposes a method and system for verifying mixed contact points based on sonar detection and localization algorithms. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, the first aspect of this invention proposes a method for verifying mixed contact points based on sonar detection and localization algorithms, comprising: S1: Obtain the first echo signal set of the target pipe section under the first working condition, extract the echo spectrum characteristics of the first echo signal set, and generate the reference echo spectrum characteristic model of the target pipe section. The first working condition is a sunny and rainless working condition or a standard working condition with known no cross-connections. S2: Obtain the second echo signal set of the target pipe section under the second operating condition, wherein the second operating condition is the operating condition that triggers the rainwater and sewage mixing event; S3: Compare and analyze the echo spectrum characteristics of the second echo signal set with the reference echo spectrum characteristic model to identify abnormal points in the target pipe section where the echo spectrum characteristics change abnormally. S4: Based on the distribution information of abnormal points, generate the verification results of the mixed connection points of the target pipe section; wherein, S4 includes: S4-1: Perform spatial clustering analysis on multiple outlier locations to generate one or more outlier location clusters; S4-2: Based on one or more clusters of anomalous points, determine the center or boundary position of each cluster of anomalous points as a candidate position for the crossover point; S4-3: Format the information of the candidate positions to generate the verification results of the mixed connection points of the target pipe segment.

[0005] Secondly, this invention proposes a hybrid contact point verification system based on sonar detection and localization algorithms. The system employs the aforementioned hybrid contact point verification method based on sonar detection and localization algorithms, and includes: The benchmark model construction module is used to perform step S1: obtain the first echo signal set of the target pipe segment under the first working condition, extract the echo spectrum features of the first echo signal set, and generate the benchmark echo spectrum feature model of the target pipe segment. The first working condition is a sunny and rainless working condition or a standard working condition with known no cross-connections. The second operating condition data acquisition module is used to execute step S2: acquire the second echo signal set of the target pipe section under the second operating condition, wherein the second operating condition is the operating condition that triggers the rainwater and sewage mixing event; The abnormal location identification module is used to perform step S3: compare and analyze the echo spectrum characteristics of the second echo signal set with the reference echo spectrum characteristic model to identify abnormal locations in the target pipe section where the echo spectrum characteristics have changed abnormally. The verification result output module is used to execute step S4: based on the distribution information of abnormal points, generate the verification results of the mixed connection points of the target pipe segment; wherein, S4 includes: S4-1: Perform spatial clustering analysis on multiple outlier locations to generate one or more outlier location clusters; S4-2: Based on one or more clusters of anomalous points, determine the center or boundary position of each cluster of anomalous points as a candidate position for the crossover point; S4-3: Format the information of the candidate positions to generate the verification results of the mixed connection points of the target pipe segment.

[0006] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention solves the technical problem mentioned in the background art of how to quickly and accurately verify the mixing point of rainwater and sewage under complex operating conditions, and realizes the precise location of the mixing point based on the change of sonar spectral characteristics. Specifically, this invention acquires the first echo signal set of the target pipe section under the first operating condition, and extracts the echo spectral characteristics to generate a reference echo spectral characteristic model. This step establishes a reliable comparison basis for subsequent anomaly identification, avoiding the subjectivity and uncertainty of traditional methods that rely on human experience. On this basis, when the second operating condition that triggers the rainwater and sewage mixing event is detected, the second echo signal set is acquired. By comparing and analyzing the echo spectral characteristics of the second echo signal set with the reference echo spectral characteristic model, the abnormal points where the echo spectral characteristics change abnormally are identified. This process makes full use of the sensitivity of sonar signals to changes in the medium inside the pipeline, and can effectively capture the spectral characteristic shift caused by the change in water flow state due to mixing, thereby accurately locating the location of the anomaly. Furthermore, this invention performs spatial clustering analysis on the identified multiple abnormal points to generate abnormal point clusters, and determines candidate locations of the crossover points based on the center or boundary positions of the clusters, finally outputting formatted crossover point verification results. This spatial clustering process effectively eliminates random errors that may be caused by single-point detection, improving the reliability and accuracy of crossover point localization.

[0007] This invention can quickly and accurately verify the connection points of rainwater and sewage under complex working conditions, thereby improving the intelligence level and detection efficiency of pipeline network operation and maintenance. Attached Figure Description

[0008] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0009] Figure 1 The diagram shown is a flowchart of a cross-connection verification method based on sonar detection and localization algorithms provided in an embodiment of the present invention. Figure 2 The diagram shown is a flowchart of the cross-connection verification method. Figure 3 The diagram shown is a schematic diagram of a cross-connection verification system based on sonar detection and localization algorithms provided in an embodiment of the present invention. Figure 4 The diagram shown illustrates the data interaction principle of the system modules. Detailed Implementation

[0010] 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. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0011] The specific embodiments of the present invention will be described below.

[0012] Example 1 like Figure 1 and Figure 2 As shown, the first aspect of this invention proposes a method for verifying mixed contact points based on sonar detection and localization algorithms, comprising: S1: Obtain the first echo signal set of the target pipe section under the first working condition, extract the echo spectrum characteristics of the first echo signal set, and generate the reference echo spectrum characteristic model of the target pipe section. The first working condition is a sunny and rainless working condition or a standard working condition with known no cross-connections. S2: Obtain the second echo signal set of the target pipe section under the second operating condition, wherein the second operating condition is the operating condition that triggers the rainwater and sewage mixing event; S3: Compare and analyze the echo spectrum characteristics of the second echo signal set with the reference echo spectrum characteristic model to identify abnormal points in the target pipe section where the echo spectrum characteristics change abnormally. S4: Based on the distribution information of abnormal points, generate the verification results of the mixed connection points of the target pipe section; wherein, S4 includes: S4-1: Perform spatial clustering analysis on multiple outlier locations to generate one or more outlier location clusters; S4-2: Based on one or more clusters of anomalous points, determine the center or boundary position of each cluster of anomalous points as a candidate position for the crossover point; S4-3: Format the information of the candidate positions to generate the verification results of the mixed connection points of the target pipe segment.

[0013] The first operating condition refers to either a clear, rain-free condition or a standard condition where no cross-contamination is known. Both conditions share the characteristic of eliminating interference from rainwater and sewage cross-contamination within the target pipe section, resulting in relatively stable water flow and minimal deposits on the pipe walls. Therefore, the acquired sonar echo signals accurately reflect the structural characteristics and media properties of the pipe section. Under clear, rain-free conditions, only normal sewage or clean water typically flows within the pipe, resulting in relatively uniform water quality. Under the standard condition where no cross-contamination is known, prior detection confirms the absence of cross-contamination points within the pipe section. Acquiring the first echo signal set under these conditions establishes a reliable benchmark for subsequent comparisons. During actual acquisition, sonar detection equipment can be deployed within the target pipe section. For example, a sonar probe can be mounted on a movable platform, moving along the pipe axis and continuously emitting sound waves while simultaneously receiving echo signals reflected from the pipe walls and the internal media. Multi-frequency sonar probes are typically used, capable of emitting sound waves of different frequencies to accommodate pipes of varying diameters and materials. The echo signal contains various components, including reflected waves from the pipe wall, scattered waves from suspended matter in the water, and reflected waves from any possible sediment. The superposition of these components constitutes a complex time-domain waveform. Through spectral analysis, such as Fast Fourier Transform, the time-domain signal can be converted into a frequency-domain distribution, yielding the amplitude or energy values ​​of different frequency components, i.e., the echo spectral characteristics. After extracting these characteristics, they are organized according to the acquisition location to form a baseline echo spectral characteristic model. This model is essentially a spectral characteristic map of various points along the pipe section under normal conditions, providing a comparative reference for subsequent anomaly identification.

[0014] Step S2 involves acquiring the second echo signal set of the target pipe section under the second operating condition. The second operating condition is defined as the condition that triggers a stormwater and sewage mixing event. For example, when online monitoring equipment in the drainage network detects a sudden increase in flow, an increase in water turbidity, or abnormal conductivity, it can be determined that a stormwater and sewage mixing event has occurred, and this method is initiated for verification. Under the second operating condition, echo signals are acquired again in the same manner as under the first operating condition, including the same detection frequency, the same sampling density, and the same travel path, to ensure the comparability of the two sets of data. The presence of mixing points causes rainwater or groundwater to enter the sewage pipes, changing the water flow velocity, water level, and water composition within the pipes. These changes directly affect the propagation path and reflection characteristics of sound waves, thus manifesting as energy changes in specific frequency bands in the echo spectrum. For example, increased water flow velocity leads to enhanced turbulence noise, and changes in water quality may alter the sound wave attenuation coefficient, all of which are reflected in the spectral characteristics. Therefore, the second echo signal set contains abnormal information caused by potential mixing points.

[0015] Step S3 involves comparing the echo spectrum characteristics of the second echo signal set with the reference echo spectrum characteristic model to identify abnormal locations in the target pipe section where the echo spectrum characteristics show unusual changes. This comparative analysis involves comparing the data collected in two sets at the same location and calculating the degree of difference in the spectral characteristics. Since the reference model represents the normal state, any difference exceeding the normal fluctuation range may indicate an anomaly. For example, the spectrum of the second echo signal can be subtracted from the spectrum at the corresponding location in the reference model to obtain the difference spectrum, and then the presence of obvious peaks or valleys can be observed. If a significant difference occurs at a certain location in a specific frequency band, that location is marked as an anomaly. Anomaly locations reflect changes in the medium or state within the pipeline and may be directly related to mixing points.

[0016] For step S4, the verification results of the mixed connection points of the target pipe section are generated based on the distribution information of the abnormal points. In S4-1, spatial clustering analysis is performed on multiple abnormal points to generate one or more abnormal point clusters. Spatial clustering analysis is used to group spatially adjacent points into the same category. Since the anomalies caused by the mixed connection points are often not isolated points, but affect a continuous area within a certain range, for example, turbulence at the mixed connection point may cause abnormal signals within a certain distance upstream and downstream. In addition, a single abnormal point may be caused by random noise or measurement error. Clustering analysis can exclude isolated points and merge consecutively occurring abnormal points into clusters. The clustering method can be based on distance, such as calculating the straight-line distance between two adjacent abnormal points. If the distance is less than a preset clustering interval threshold, they are grouped into the same cluster. The abnormal point clusters represent continuous sections where mixed connections may exist.

[0017] In S4-2, based on one or more clusters of anomalous points, the center or boundary location of each cluster is determined as a candidate location for the cross-cutting point. The center location can be the geometric mean of all anomalous points within the cluster, reflecting the center of the anomalous area; the boundary location can be the two endpoints or outer edge of the area covered by the cluster. Choosing the center location as a candidate is suitable for cases where the anomalous points are symmetrically distributed; choosing the boundary location as a candidate is suitable for cases where the anomalous points spread from a certain point to one side. Transforming discrete anomalous points into specific candidate locations facilitates subsequent on-site verification or repair.

[0018] In S4-3, the information of candidate locations is formatted to generate verification results for the mixed connection points of the target pipe segment. Formatting may include converting location coordinates to a standard geographic coordinate format, adding confidence levels, generating visualizations, or exporting data reports. For example, candidate locations can be marked on the pipeline longitudinal profile, along with corresponding spectral difference values ​​and the number of outliers, forming an intuitive verification report. Such results can be directly provided to maintenance personnel as a basis for accurate repairs.

[0019] First, the establishment of the baseline model provides a reference standard for anomaly identification; second, data acquisition under the second operating condition ensures the real-time nature of the information; third, spectral feature comparison accurately captures the physical changes caused by the misconnection; finally, spatial clustering analysis eliminates random interference and focuses the anomaly on specific candidate locations. This invention utilizes the sensitivity of sonar spectral features to changes in the internal state of the pipeline, combined with a spatial clustering algorithm, to effectively solve the problem of traditional methods' difficulty in accurately locating misconnection points, thus improving the accuracy and reliability of detection.

[0020] In some implementations, S1 includes: S1-1: Control the mobile inspection platform equipped with a sonar probe to move within the target pipe section according to a preset path and generate real-time location information of the mobile inspection platform; S1-2: Based on the movement of the mobile detection platform, control the sonar probe to emit the first detection sound wave towards the pipe wall, and receive the first echo signal reflected from the pipe wall, and record and generate the first echo signal set; S1-3: Perform spectral decomposition on each echo signal in the first echo signal set to extract and generate reference spectral features associated with the physical state of the pipe wall; S1-4: Associate and store the reference spectrum features with the real-time location information to generate a reference echo spectrum feature model of the target pipe section.

[0021] In step S1-1, a mobile detection platform equipped with a sonar probe is controlled to travel along a preset path within the target pipe section, generating real-time position information for the mobile detection platform. The mobile detection platform refers to a device capable of autonomous movement underwater or within a pipeline, such as an underwater unmanned vehicle or a pipe-crawling robot. Its sonar probe can be a multi-frequency sonar probe capable of emitting and receiving sound waves. The preset path needs to cover the entire target pipe section, typically along the pipe axis from upstream to downstream or vice versa, ensuring no blind spots. The travel speed needs to match the sampling frequency of the sonar probe to ensure sufficient overlap between adjacent sampling points and avoid missed detections. Real-time position information can be obtained in various ways, such as installing a mileage wheel on the pipe-crawling robot to calculate the travel distance based on the wheel's rotation count; or installing an inertial navigation system on the underwater unmanned vehicle to calculate the distance using the absolute coordinates at the pipe inlet. The position information is recorded synchronously with a timestamp as a basis for subsequent spatial correlation.

[0022] In steps S1-2, based on the movement of the mobile detection platform, the sonar probe is controlled to emit a first detection sound wave towards the pipe wall and receive the first echo signal reflected from the pipe wall, recording and generating a first echo signal set. The frequency of the sound wave emitted by the sonar probe can be selected according to the pipe material and diameter. For example, a lower frequency can be used for concrete pipes to reduce attenuation, while a higher frequency can be used for metal pipes to improve resolution. The emitted sound wave is sent in pulse form, and after encountering the pipe wall or the medium inside the pipe, it is reflected back and received by the same probe. The received echo signal is amplified, filtered, and converted from analog to digital before being stored in digital form. During storage, each echo signal corresponds to a transmission time, and the real-time position information at that time is recorded simultaneously, forming a data record containing position and waveform. The first echo signal set generated in this way is a series of waveform data distributed along the pipe.

[0023] In steps S1-3, spectral decomposition is performed on each echo signal in the first echo signal set to extract and generate reference spectral features associated with the physical state of the pipe wall. Spectral decomposition typically employs a Fast Fourier Transform (FFT) algorithm to transform each echo signal from the time domain to the frequency domain, obtaining an amplitude spectrum or power spectrum. The physical state of the pipe wall includes its material, thickness, smoothness of the inner wall, presence of cracks or corrosion, and the presence of deposits on the pipe wall. These conditions affect the reflection coefficient and scattering characteristics of the sound waves, thus forming specific patterns in the spectrum. For example, a smooth metal pipe wall produces strong specular reflection, manifested as a narrow peak in the spectrum; a rough concrete pipe wall produces diffuse reflection, with a wider spectrum and dispersed energy; if there are soft deposits on the pipe wall, they will absorb some high-frequency components, leading to a decrease in high-frequency energy. By analyzing the shape, peak frequency, bandwidth, and energy distribution of the spectrum, quantitative features characterizing these physical states can be extracted. Correlating these features with their corresponding locations forms the reference spectral features.

[0024] In steps S1-4, the reference spectral features are associated with and stored in real-time location information to generate a reference echo spectral feature model of the target pipe segment. Associated storage means storing the spectral features of each location point as an attribute of that location in a database or file. For example, a two-dimensional table can be created, with each row corresponding to a sampling point, containing location coordinates and a set of spectral feature vectors. This model is equivalent to the spectral fingerprint of the pipe segment under normal operating conditions, providing an accurate reference benchmark for subsequent anomaly comparison.

[0025] The autonomous movement of the mobile detection platform ensures the comprehensiveness and consistency of data collection, avoiding omissions that may occur during manual operation; the features extracted by spectral decomposition have clear physical meanings and can reflect subtle changes in the pipe wall state; the positional correlation gives the model spatial resolution capabilities, accurate to each sampling point; the establishment of the benchmark model provides a reliable comparison standard for subsequent anomaly identification.

[0026] In some implementations, S2 includes: S2-1: Based on the second working condition, control the mobile inspection platform to enter the target pipe section again, and travel along the same preset path as in S1-1, generating real-time location information of the mobile inspection platform's re-travel. S2-2: Based on the re-travel process of the mobile detection platform, control the sonar probe to emit a second detection sound wave towards the pipe wall, and receive the second echo signal reflected from the pipe wall, and record and generate the second echo signal set.

[0027] In step S2-1, based on the second operating condition, the mobile inspection platform is again controlled to enter the target pipe section and travel along the same preset path as in step S1-1, generating real-time position information of the mobile inspection platform's second movement. The same preset path ensures that the spatial sampling points collected in the two trials are as consistent as possible, thus ensuring accurate positional correspondence during subsequent comparisons. In actual operation, the mobile inspection platform can be programmed to automatically repeat the previous trajectory, or it can be remotely controlled to move along the same route. The real-time position information of the second movement also needs to be accurately recorded, using the same method as the first time, such as odometer, inertial navigation, or external positioning. Since the pipeline environment may change over time, such as increased sediment or changes in water flow velocity, the consistency of the path minimizes comparison errors caused by positional deviations, ensuring that subsequent difference analysis primarily reflects changes in operating conditions rather than positional biases.

[0028] In step S2-2, based on the re-movement of the mobile detection platform, the sonar probe is controlled to emit a second detection sound wave towards the pipe wall and receive the second echo signal reflected from the pipe wall, recording and generating a second echo signal set. The emission parameters of the second detection sound wave, including emission frequency, pulse width, emission power, and sampling frequency, should be exactly the same as the first detection sound wave to eliminate systematic errors caused by different detection conditions. The received second echo signal is also recorded synchronously with the real-time position information of the re-movement, forming a second echo signal set. Through this process, an echo dataset under the mixed-interaction event triggering condition is obtained. This dataset has the same spatial sampling points, the same detection parameters, and the same sampling density as the benchmark dataset under the first condition, and therefore is directly comparable.

[0029] This method ensures that the only variable between two sets of data is the pipeline condition itself by maintaining consistency between the travel path and the detection parameters, thus enabling subsequent anomaly identification to focus on changes caused by misconnection.

[0030] In some implementations, S3 includes: S3-1: Based on the real-time position information of the second echo signal, perform time synchronization or spatial registration processing on each second echo signal in the second echo signal set and the corresponding reference spectrum feature in the reference echo spectrum feature model to generate the registered second echo signal and the registered reference spectrum feature. S3-2: Compare the registered second echo signal with the registered reference spectrum characteristics point by point, and calculate the difference in spectrum characteristics within the preset frequency band; S3-3: Mark the locations where the spectral feature difference value exceeds the preset change threshold to generate anomaly points.

[0031] In step S3-1, based on the real-time position information of the second echo signal during re-travel, each second echo signal in the second echo signal set is time-synchronized or spatially registered with the corresponding reference spectrum feature in the reference echo spectrum feature model to generate the registered second echo signal and the registered reference spectrum feature. Since the mobile detection platform may experience slight speed fluctuations or positioning deviations during two journeys, the sampling points corresponding to the same physical location may not completely overlap. For example, the first sampling might be at location A, while the second sampling might be at location A', which is near location A. Therefore, registration is necessary for correction. Time synchronization refers to alignment based on a timestamp, suitable for situations where the travel speed is constant. Spatial registration utilizes position information for interpolation or matching, for example, using linear interpolation to find the reference feature closest to the second echo signal's position in the reference model. In practice, the real-time position information of the second echo signal during re-travel can be used as a query key to find the nearest location point in the reference model. If the distance is less than the allowable error, a direct match is performed; if the distance is large, interpolation is performed on nearby reference spectrum features to generate estimated features for the corresponding location. The second echo signal generated in this way corresponds precisely in space to the registered reference spectrum characteristics, ensuring that subsequent comparisons are of signals from the same physical point.

[0032] In step S3-2, the registered second echo signal is compared point-by-point with the registered reference spectral characteristics, and the spectral characteristic difference value within the preset frequency band is calculated. Point-by-point comparison means comparing the spectral characteristics of the second echo signal with the reference spectral characteristics at each location point. The preset frequency band can be pre-selected based on the pipe characteristics and mixing sensitivity, for example, selecting a frequency band near the center frequency of the sonar probe because this band has a high signal-to-noise ratio; or selecting a frequency band sensitive to mixing, for example, based on previous experiments showing that mixing causes energy changes in a specific frequency band. The method for calculating the difference value can take various forms, such as calculating the absolute value of the difference in energy values ​​between the two spectra within the preset frequency band, or calculating the sum of squares of the spectral amplitude differences, or calculating the correlation coefficient between the two spectra. Taking the energy difference as an example, firstly, the registered second echo signal and the registered reference spectral characteristics are subjected to Fast Fourier Transform to obtain their respective spectral distribution maps; then, the spectral distribution maps are integrated or summed within the preset frequency band to obtain their respective spectral energy values; finally, the absolute value of the difference between these two energy values ​​is calculated as the spectral characteristic difference value. This difference value quantifies the degree of change of that location within a preset frequency band.

[0033] In step S3-3, locations where the spectral feature difference value exceeds a preset change threshold are marked as abnormal points. When the difference value exceeds the threshold, it indicates that a significant change beyond normal fluctuations has occurred at that location, potentially indicating the existence of a mixing point. Through this marking, a series of abnormal points are obtained, which may appear continuously in space or exist in isolation. This invention, through spectral analysis and frequency band selection, can capture subtle changes caused by medium variations. For example, water flow turbulence caused by mixing may produce energy enhancement in the high-frequency band, while water quality changes may cause attenuation in the mid-frequency band. These changes manifest as specific patterns in the spectrum, which can be quantified through difference value calculation. Simultaneously, the threshold judgment is based on statistical or experimental data and has a certain degree of adaptability, capable of adapting to the normal fluctuation levels of different pipe sections, thereby improving the accuracy and generalization ability of anomaly identification.

[0034] In some implementations, S3-2 includes: Fast Fourier Transform is performed on the registered second echo signal and the registered reference spectral characteristics to generate corresponding spectral distribution maps. The spectral energy value is calculated on the spectral distribution map to generate the absolute value of the difference between the spectral energy values ​​corresponding to the registered second echo signal and the registered reference spectral characteristics within the preset frequency band, which is used as the spectral characteristic difference value.

[0035] In a specific implementation, after spatial registration of the second echo signal and the reference spectral features is completed, signal processing techniques are needed to transform and compare the registered signals in order to quantify their spectral differences. Specifically, the registered second echo signal and the registered reference spectral features are processed using Fast Fourier Transform (FFT). FFT converts a time-domain signal into a frequency-domain signal, revealing the different frequency components and their amplitudes contained in the signal. This transformation generates a spectral distribution map corresponding to the registered second echo signal and a spectral distribution map corresponding to the registered reference spectral features. The spectral distribution map typically uses frequency as the x-axis and amplitude or power as the y-axis to visually display the energy distribution of the signal at each frequency point. In practice, the number of FFT points must balance frequency resolution and computational efficiency, usually using a number of transform points that matches the signal length to ensure spectral precision. For example, if the sampling frequency is constant, more transform points result in higher frequency resolution, but also increase the computational load. Therefore, a trade-off must be made based on actual hardware conditions and detection requirements. The resulting spectrum distribution can be either an amplitude spectrum or a power spectrum. The amplitude spectrum reflects the magnitude of each frequency component, while the power spectrum reflects the energy distribution of each frequency component. Both can be used for subsequent energy calculations.

[0036] Based on this, the spectral energy value is calculated from the generated spectral distribution map. The spectral energy value can be defined as the integral or summation of the square of the spectral amplitude within a specific frequency band, reflecting the total energy of the signal within that band. Preset frequency bands can be selected in advance according to the actual application scenario; for example, the main frequency band within the sonar probe's operating frequency range can be selected, or a characteristic frequency band sensitive to mixing events can be selected. When selecting preset frequency bands, it is necessary to analyze the pipe material, pipe diameter, medium characteristics, and historical data to determine which frequency bands are most sensitive to medium changes caused by mixing. For example, mixing may lead to enhanced water turbulence, thereby generating additional noise energy in the high-frequency band; or the sediment deposition caused by mixing may absorb sound waves of specific frequencies, leading to energy attenuation in that frequency band. Therefore, the preset frequency band can be a continuous frequency range or multiple discontinuous frequency bands, each corresponding to different physical mechanisms. Through calculation, the spectral energy value of the registered second echo signal within the preset frequency band and the spectral energy value of the registered reference spectral characteristics within the same preset frequency band are obtained. Then, the absolute value of the difference between these two spectral energy values ​​is calculated, and this absolute value is used as the spectral characteristic difference value. This difference value quantifies the degree of energy change at that location point relative to the baseline state within a preset frequency band under the second operating condition. A large difference value indicates a significant change in the echo spectrum energy at that location point, potentially foreshadowing changes in the medium or pipe wall condition caused by misconnection. This method allows for an objective and quantitative assessment of the anomaly degree at each location point, avoiding the subjectivity of manual interpretation.

[0037] This method can capture subtle frequency domain changes caused by cross-connection. For example, water turbulence caused by cross-connection may generate energy enhancement in the high-frequency band, while sediment changes may cause attenuation in the mid-frequency band. These changes can be effectively detected through energy differences. Therefore, this embodiment further improves the sensitivity and accuracy of anomaly identification, making cross-connection point verification more scientific and objective.

[0038] In some implementations, S4-1 includes: Calculate the straight-line distance between two adjacent anomaly points; If the straight-line distance is less than the preset clustering distance threshold, then two adjacent outlier points will be grouped into the same outlier point cluster, generating one or more outlier point clusters.

[0039] After identifying multiple anomaly points, these points may be scattered across different locations within the target pipe section. Some are continuous anomaly areas caused by actual cross-connections, while others may be isolated noise or measurement errors. To distinguish between these two situations, spatial clustering analysis of the anomaly points is required. Specifically, the straight-line distance between two adjacent anomaly points is first calculated. Here, "adjacent" refers to two anomaly points arranged sequentially along the pipe's axial direction. The straight-line distance is the projected distance between the two points along the pipe's axis. If the pipe is a straight segment, the Euclidean distance is directly calculated; if the pipe is curved, the curvilinear distance can be calculated along the pipe's centerline. However, for simplicity, the straight-line distance can be approximated because the distance between adjacent points is usually small, and the influence of pipe curvature can be ignored. In actual calculations, the precise coordinates of each anomaly point are obtained based on the real-time location information recorded by the mobile detection platform. Then, the points are sorted according to the direction of travel, and the spatial distance between adjacent points is calculated sequentially. Finally, this straight-line distance is compared with a preset clustering interval threshold. The size of the clustering interval threshold can be determined based on factors such as the pipe diameter, the sampling interval of the detection platform, and the potential impact range of the cross-connection points. For example, if the sampling interval is small and the distance between adjacent sampling points is small, anomalies caused by aliasing may cover multiple consecutive sampling points. Therefore, the threshold can be set slightly larger than the sampling interval to merge consecutive points. If the sampling interval is large, the actual spatial range of the aliasing effect needs to be considered, and the threshold needs to be set larger. If the straight-line distance between two adjacent anomaly points is less than the preset clustering distance threshold, these two anomaly points are considered to be spatially close enough and may belong to the same aliasing event-induced anomaly region, so they are grouped into the same anomaly point cluster. This process is repeated for all anomaly points, gradually merging adjacent points that meet the distance condition to ultimately generate one or more anomaly point clusters. Each anomaly point cluster contains several spatially continuous or adjacent anomaly points, reflecting potentially aliased continuous segments. If the distance between some anomaly points and their preceding and following points is greater than the threshold, they may form isolated points. These isolated points can be excluded as noise or marked separately for further analysis. This distance-based clustering method effectively integrates scattered anomalies into meaningful clusters, preserving spatial information while reducing random interference.

[0040] This method generates clusters through clustering and then determines candidate locations based on these clusters, thus avoiding single-point errors and improving the reliability of the verification results. After clustering, the area covered by each anomalous point cluster is the suspected cross-connection segment. Subsequent steps will further determine candidate locations based on this cluster. This embodiment provides an effective spatial filtering method for locating cross-connection points.

[0041] In some implementations, the mobile detection platform is an underwater unmanned vehicle or a pipeline crawling robot, and the sonar probe is a multi-frequency sonar probe.

[0042] Mobile inspection platforms can be specifically implemented as underwater unmanned vehicles (UAVs) or pipe-crawling robots, while sonar probes can be specifically implemented as multi-frequency sonar probes. Underwater UAVs are devices capable of autonomous or remotely controlled navigation in water, typically equipped with propulsion systems, navigation systems, and various sensors. When applied to drainage pipe inspection, underwater UAVs can submerge into water-filled pipes and travel along the pipe axis. Their flexibility and maneuverability allow them to adapt to the inspection needs of pipes of different diameters and complex structures. Underwater UAVs are usually powered by batteries and driven by propellers or water jets. They utilize built-in inertial navigation systems, depth gauges, and sonar positioning systems to achieve autonomous navigation and position recording. They can travel continuously for several kilometers inside pipes, collecting sonar echo data along the way. Pipe-crawling robots are robots specifically designed to crawl along the inner walls of pipes. They are typically wheeled or tracked and can adhere to the pipe walls for movement. They are suitable for dry or shallow water environments and can also work in pipes with a certain water depth. Pipe-crawling robots are typically powered and controlled by cables, enabling remote real-time video monitoring and sensor data acquisition. Their controllable speed and precise positioning make them suitable for applications requiring detailed inspection. Both platforms have their advantages, and the choice can be made based on the actual conditions of the target pipe section. Underwater unmanned vehicles are suitable for water-filled pipes, suspending themselves in the water for detection, reducing friction with the pipe walls and unaffected by the degree of siltation within the pipe; pipe-crawling robots are suitable for scenarios with low water levels or where close contact with the pipe wall is required for inspection, offering more stable signal acquisition and ease of mounting various sensors.

[0043] The sonar probe employs multi-frequency sonar, meaning it can emit and receive sound waves of multiple frequencies. By emitting sound waves at multiple frequencies, richer echo information can be obtained, reflecting the pipe's condition from different angles. For example, the low-frequency channel can be used to detect cavities behind the pipe wall or external interference, while the high-frequency channel can be used to identify cracks or deposits in the pipe wall. A multi-frequency sonar probe can be implemented by integrating multiple transducer units within a single probe, each operating at a different frequency; or it can be a single broadband transducer that emits broadband pulses and uses a receiving filter to separate different frequency components. The received signals are amplified, filtered, and converted from analog to digital, then stored as data streams corresponding to different frequencies. In subsequent processing, spectral analysis can be performed on the signals at each frequency separately, or the multi-frequency information can be fused into a comprehensive feature.

[0044] In practice, the appropriate underwater unmanned vehicle or pipeline crawling robot model can be selected based on the water flow velocity, turbidity, water depth and detection accuracy requirements in the pipeline, and the corresponding multi-frequency sonar probe can be configured to ensure that the collected echo signal has sufficient quality and information content.

[0045] Example 2 like Figure 3 and Figure 4 As shown, in a second aspect, the present invention proposes a hybrid contact point verification system based on sonar detection and localization algorithms. The system employs the hybrid contact point verification method based on sonar detection and localization algorithms proposed in any of the above embodiments. The system includes: The benchmark model construction module is used to perform step S1: obtain the first echo signal set of the target pipe segment under the first working condition, extract the echo spectrum features of the first echo signal set, and generate the benchmark echo spectrum feature model of the target pipe segment. The first working condition is a sunny and rainless working condition or a standard working condition with known no cross-connections. The second operating condition data acquisition module is used to execute step S2: acquire the second echo signal set of the target pipe section under the second operating condition, wherein the second operating condition is the operating condition that triggers the rainwater and sewage mixing event; The abnormal location identification module is used to perform step S3: compare and analyze the echo spectrum characteristics of the second echo signal set with the reference echo spectrum characteristic model to identify abnormal locations in the target pipe section where the echo spectrum characteristics have changed abnormally. The verification result output module is used to execute step S4: based on the distribution information of abnormal points, generate the verification results of the mixed connection points of the target pipe segment; wherein, S4 includes: S4-1: Perform spatial clustering analysis on multiple outlier locations to generate one or more outlier location clusters; S4-2: Based on one or more clusters of anomalous points, determine the center or boundary position of each cluster of anomalous points as a candidate position for the crossover point; S4-3: Format the information of the candidate positions to generate the verification results of the mixed connection points of the target pipe segment.

[0046] This system corresponds to the method proposed in Example 1, and will not be described in detail here.

[0047] 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 or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for verifying cross-connection points based on sonar detection and localization algorithms, characterized in that, include: S1: Obtain the first echo signal set of the target pipe section under the first working condition, extract the echo spectrum characteristics of the first echo signal set, and generate the reference echo spectrum characteristic model of the target pipe section. The first working condition is a sunny and rainless working condition or a standard working condition with known no cross-connections. S2: Obtain the second echo signal set of the target pipe section under the second operating condition, wherein the second operating condition is the operating condition that triggers the rainwater and sewage mixing event; S3: Compare and analyze the echo spectrum characteristics of the second echo signal set with the reference echo spectrum characteristic model to identify abnormal points in the target pipe section where the echo spectrum characteristics change abnormally. S4: Based on the distribution information of abnormal points, generate the verification results of the mixed connection points of the target pipe section; wherein, S4 includes: S4-1: Perform spatial clustering analysis on multiple outlier locations to generate one or more outlier location clusters; S4-2: Based on one or more clusters of anomalous points, determine the center or boundary position of each cluster of anomalous points as a candidate position for the crossover point; S4-3: Format the information of the candidate positions to generate the verification results of the mixed connection points of the target pipe segment; S1 includes: S1-1: Control the mobile inspection platform equipped with a sonar probe to move within the target pipe section according to a preset path and generate real-time location information of the mobile inspection platform; S1-2: Based on the movement of the mobile detection platform, control the sonar probe to emit the first detection sound wave towards the pipe wall, and receive the first echo signal reflected from the pipe wall, and record and generate the first echo signal set; S1-3: Perform spectral decomposition on each echo signal in the first echo signal set to extract and generate reference spectral features associated with the physical state of the pipe wall; S1-4: Associate and store the reference spectrum features with the real-time location information to generate a reference echo spectrum feature model of the target pipe segment; S2 includes: S2-1: Based on the second working condition, control the mobile inspection platform to enter the target pipe section again, and travel along the same preset path as in S1-1, generating real-time location information of the mobile inspection platform's re-travel. S2-2: Based on the re-travel process of the mobile detection platform, control the sonar probe to emit a second detection sound wave towards the pipe wall, and receive the second echo signal reflected from the pipe wall, and record and generate the second echo signal set.

2. The method for verifying mixed contact points based on sonar detection and localization algorithms according to claim 1, characterized in that, S3 include: S3-1: Based on the real-time position information of the second echo signal, perform time synchronization or spatial registration processing on each second echo signal in the second echo signal set and the corresponding reference spectrum feature in the reference echo spectrum feature model to generate the registered second echo signal and the registered reference spectrum feature. S3-2: Compare the registered second echo signal with the registered reference spectrum characteristics point by point, and calculate the difference in spectrum characteristics within the preset frequency band; S3-3: Mark the locations where the spectral feature difference value exceeds the preset change threshold to generate anomaly points.

3. The method for verifying mixed contact points based on sonar detection and localization algorithms according to claim 2, characterized in that, S3-2 includes: Fast Fourier Transform is performed on the registered second echo signal and the registered reference spectral characteristics to generate corresponding spectral distribution maps. The spectral energy value is calculated on the spectral distribution map to generate the absolute value of the difference between the spectral energy values ​​corresponding to the registered second echo signal and the registered reference spectral characteristics within the preset frequency band, which is used as the spectral characteristic difference value.

4. The method for verifying mixed contact points based on sonar detection and localization algorithms according to claim 1, characterized in that, S4-1 includes: Calculate the straight-line distance between two adjacent anomaly points; If the straight-line distance is less than the preset clustering distance threshold, then two adjacent outlier points will be grouped into the same outlier point cluster, generating one or more outlier point clusters.

5. The method for verifying mixed contact points based on sonar detection and localization algorithms according to claim 1, characterized in that, The mobile detection platform is an underwater unmanned vehicle or a pipeline crawling robot, and the sonar probe is a multi-frequency sonar probe.

6. A hybrid contact point verification system based on sonar detection and localization algorithms, characterized in that, The system employs the hybrid contact point verification method based on sonar detection and localization algorithms as described in any one of claims 1 to 5, and the system comprises: The benchmark model construction module is used to perform step S1: obtain the first echo signal set of the target pipe segment under the first working condition, extract the echo spectrum features of the first echo signal set, and generate the benchmark echo spectrum feature model of the target pipe segment. The first working condition is a sunny and rainless working condition or a standard working condition with known no cross-connections. The second operating condition data acquisition module is used to execute step S2: acquire the second echo signal set of the target pipe section under the second operating condition, wherein the second operating condition is the operating condition that triggers the rainwater and sewage mixing event; The abnormal location identification module is used to perform step S3: compare and analyze the echo spectrum characteristics of the second echo signal set with the reference echo spectrum characteristic model to identify abnormal locations in the target pipe section where the echo spectrum characteristics have changed abnormally. The verification result output module is used to execute step S4: based on the distribution information of abnormal points, generate the verification results of the mixed connection points of the target pipe segment; wherein, S4 includes: S4-1: Perform spatial clustering analysis on multiple outlier locations to generate one or more outlier location clusters; S4-2: Based on one or more clusters of anomalous points, determine the center or boundary position of each cluster of anomalous points as a candidate position for the crossover point; S4-3: Format the information of the candidate positions to generate the verification results of the mixed connection points of the target pipe segment.

Images