Water supply pipeline leakage detection method and device, electronic equipment and storage medium

Abstract

The invention belongs to the technical field of buried pipe network detection, and particularly discloses a water supply pipeline leakage detection method and device, electronic equipment and a storage medium. The method comprises the steps of determining suspected leakage points along a to-be-detected pipeline based on temperature information along the to-be-detected pipeline; determining time-frequency characteristic information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point; inputting the time-frequency characteristic information of the multi-source monitoring information into a pipeline leakage detection model to obtain leakage state information of the suspected leakage point output by the pipeline leakage detection model; the multi-source monitoring information comprises temperature information, vibration information and strain information; the leakage detection model is obtained by training according to time-frequency characteristic information of a multi-source monitoring information sample and a corresponding pipeline leakage state label. Through the pipeline leakage detection method and device, the detection efficiency can be improved while the pipeline leakage detection precision and reliability can be effectively improved, and the labor cost is reduced.

Description

Technical Field

[0001] The present application belongs to the technical field of pipe network detection, and more specifically, relates to a water supply pipe leakage detection method, device, electronic equipment and storage medium. Background Art

[0002] Buried water supply pipelines often face pipeline leakage problems during operation, which not only leads to waste of water resources but also may affect water supply safety.

[0003] Traditional water supply pipeline leakage detection methods, such as manual inspections or manual leak detection using leak detection rods, lack real-time detection and accuracy, and require a lot of manpower, resulting in high labor costs. With the development of sensor technology, the emergence of distributed fiber optic sensing technology has brought new opportunities for pipeline leakage detection. Although distributed fiber optic sensing technology has been applied to certain projects in this type of monitoring field, due to the limited consideration of influencing factors and the single detection data type, this method currently has problems such as low detection accuracy and insufficient reliability.

[0004] Therefore, how to better detect water supply pipeline leakage has become a technical problem that needs to be solved urgently in the industry. Summary of the invention

[0005] In view of the defects of the prior art, the purpose of this application is to better realize the detection of water supply pipe leakage, aiming to solve the problems of low detection accuracy and insufficient reliability of existing detection methods.

[0006] To achieve the above objectives, in a first aspect, the present application provides a water supply pipeline leakage detection method, comprising: Determine suspected leakage points along the pipeline to be tested based on temperature information along the pipeline to be tested; Based on the multi-source monitoring information of the suspected leakage point, determining the time-frequency characteristic information of the multi-source monitoring information; Inputting the time-frequency characteristic information of the multi-source monitoring information into a pipeline leakage detection model to obtain leakage status information of the suspected leakage point output by the pipeline leakage detection model; The multi-source monitoring information includes temperature information, vibration information and strain information; the leakage detection model is trained based on the time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage state labels.

[0007] Optionally, the determining the time-frequency feature information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point includes: Determine, according to the temperature information of the suspected leakage point, multiple types of time domain feature information corresponding to the temperature information; Determine, according to the vibration information of the suspected leakage point, multiple types of time-frequency feature information corresponding to the vibration information; Determine multiple types of time-domain feature information corresponding to the strain information according to the strain information of the suspected leakage point; Based on the multiple types of time-domain feature information corresponding to the temperature information, the multiple types of time-frequency feature information corresponding to the vibration information, and the multiple types of time-domain feature information corresponding to the strain information, determine the time-frequency feature information of the multi-source monitoring information.

[0008] Optionally, after determining the multiple types of time-domain feature information corresponding to the strain information according to the strain information of the suspected leakage point, the method further includes: Determine the feature information of multiple types of first target features corresponding to the temperature information from the multiple types of time-domain feature information corresponding to the temperature information; Determine the feature information of multiple types of second target features corresponding to the vibration information from the multiple types of time-frequency feature information corresponding to the vibration information; Determine the feature information of multiple types of third target features corresponding to the strain information from the multiple types of time-domain feature information corresponding to the strain information; the multiple types of first target features, the multiple types of second target features, and the multiple types of third target features are determined by performing time-frequency feature screening based on multi-source monitoring information samples and corresponding pipeline leakage status labels; Based on the feature information of each of the first target features, the feature information of each of the second target features, and the feature information of each of the third target features, determine the time-frequency feature information of the multi-source monitoring information.

[0009] Optionally, before determining the time-frequency feature information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point, the method further includes: According to the multi-source monitoring information samples corresponding to each pipeline leakage status label, determine multiple types of time-domain feature information of the temperature information samples, multiple types of time-frequency feature information of the vibration information samples, and multiple types of time-domain feature information of the strain information samples corresponding to each pipeline leakage status label; For any feature information among the multiple types of time-domain feature information of the temperature information samples, the multiple types of time-frequency feature information of the vibration information samples, and the multiple types of time-domain feature information of the strain information samples, determine the Minkowski distance between any feature information under different pipeline leakage states; According to the Minkowski distance between any feature information under different pipeline leakage states, determine whether the feature corresponding to the any feature information is a fourth target feature, so as to determine multiple types of the fourth target features and the remaining features other than the multiple types of the fourth target features; Perform rank correlation coefficient analysis according to the feature information of the remaining features, and determine multiple types of fifth target features from the remaining features; Determine multiple categories of the first target feature, multiple categories of the second target feature, and multiple categories of the third target feature according to multiple categories of the fourth target feature and multiple categories of the fifth target feature.

[0010] Optionally, the multiple categories of time-domain feature information corresponding to the temperature information include maximum value, average value, time-domain range, standard deviation, third-order origin moment, third-order central moment, skewness, and kurtosis; The multiple categories of time-frequency feature information corresponding to the vibration information include time-domain range, standard deviation, spectral standard deviation, spectral skewness, spectral kurtosis, spectral entropy, strongest frequency, and ratio of main peak to secondary peak; The multiple categories of time-domain feature information corresponding to the strain information include maximum value, minimum value, average value, time-domain range, standard deviation, root mean square, waveform index, skewness, kurtosis, impulse index, and zero-crossing rate.

[0011] Optionally, before inputting the time-frequency feature information of the multi-source monitoring information into the pipeline leakage detection model to obtain the leakage status information of the suspected leakage point output by the pipeline leakage detection model, the method further includes: Taking the time-frequency feature information of the multi-source monitoring information sample and the corresponding pipeline leakage status label as a set of training samples, and obtaining multiple sets of the training samples; Training a random forest model with multiple sets of the training samples to obtain a trained random forest model, and using the trained random forest model as the pipeline leakage detection model.

[0012] In a second aspect, the present application provides a water supply pipeline leakage detection device, including: A first processing module, configured to determine a suspected leakage point along the pipeline to be measured based on the temperature information along the pipeline to be measured; A second processing module, configured to determine the time-frequency feature information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point; A leakage detection module, configured to input the time-frequency feature information of the multi-source monitoring information into the pipeline leakage detection model to obtain the leakage status information of the suspected leakage point output by the pipeline leakage detection model; The multi-source monitoring information includes temperature information, vibration information, and strain information; the leakage detection model is trained according to the time-frequency feature information of the multi-source monitoring information sample and the corresponding pipeline leakage status label.

[0013] In a third aspect, the present application provides an electronic device, including: at least one memory for storing a program; at least one processor for executing the program stored in the memory, and when the program stored in the memory is executed, the processor is used to execute the method described in the first aspect or any possible implementation manner of the first aspect.

[0014] Fourthly, the present application provides a computer-readable storage medium storing a computer program, which, when running on a processor, causes the processor to execute the method described in the first aspect or any possible implementation manner of the first aspect.

[0015] Fifthly, the present application provides a computer program product, which, when running on a processor, causes the processor to execute the method described in the first aspect or any possible implementation manner of the first aspect.

[0016] It can be understood that the beneficial effects of the above-mentioned second aspect to fifth aspect can refer to the relevant descriptions in the first aspect above, and will not be elaborated here.

[0017] Generally speaking, compared with the prior art, the above technical solution conceived by the present application has the following beneficial effects: A water supply pipeline leakage detection method, device, electronic device and storage medium provided by the present application, by integrating the advantages of various monitoring signals, considering the internal influence relationship between pipeline leakage and each monitoring signal, after locking the suspected leakage points along the pipeline to be measured, using multi-source monitoring information including temperature information, vibration information and strain information, and fusing various time-frequency characteristics of the multi-source monitoring information to more comprehensively detect and identify the leakage state of the suspected leakage points, without the need for a large amount of manual participation, can effectively improve the pipeline leakage detection accuracy and reliability while improving the detection efficiency and reducing the labor cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is one of the schematic flowcharts of the water supply pipeline leakage detection method provided by the embodiment of the present application; Figure 2 is the second schematic flowchart of the water supply pipeline leakage detection method provided by the embodiment of the present application; Figure 3 is the schematic flowchart of the water supply pipeline leakage condition assessment provided by the embodiment of the present application; Figure 4 The structural schematic diagram of the water supply pipeline leakage detection device provided by the embodiment of the present application; Figure 5 is the structural schematic diagram of the electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0020] In this application, the term "and / or" describes the relationship between related objects and indicates that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0021] In the description of the specification and claims of this application, terms such as "first" and "second" are used to distinguish different objects, rather than to describe the specific order of the objects. For example, the first target feature and the second target feature are used to distinguish the time-domain features or time-frequency features of different detection signals, rather than to describe the specific order of the target features.

[0022] In the embodiments of this application, words such as "exemplary" or "for example" are used to represent examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner.

[0023] In the description of the embodiments of this application, unless otherwise specified, the meaning of "multiple categories" refers to two or more categories. For example, multiple categories of time-domain feature information refer to two or more categories of time-domain feature information, etc.

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

[0025] In the early stage or when the leakage of the water supply pipeline is small, it may be missed due to the tiny changes in temperature, vibration, and strain, and it is also prone to false alarms due to interference from external factors. For example, the temperature signal is affected by sunlight and climate, the vibration signal is affected by vehicle driving and machine operation, and the strain signal is affected by geological conditions. Therefore, it is necessary to combine multiple distributed optical fiber sensing technologies to achieve leakage detection of the water supply pipeline through multi-source monitoring signal data fusion.

[0026] Figure 1 is one of the schematic flowcharts of the water supply pipeline leakage detection method provided by the embodiments of this application, as Figure 1 shown, including: Step S1, based on the temperature information along the pipeline to be measured, determine the suspected leakage points along the pipeline to be measured; Step S2, based on the multi-source monitoring information of the suspected leakage points, determine the time-frequency feature information of the multi-source monitoring information; Step S3, input the time-frequency feature information of the multi-source monitoring information into the pipeline leakage detection model to obtain the leakage status information of the suspected leakage points output by the pipeline leakage detection model; The multi-source monitoring information includes temperature information, vibration information, and strain information; the leakage detection model is trained based on the time-frequency characteristic information of the multi-source monitoring information samples and the corresponding pipeline leakage status labels.

[0027] Specifically, the suspected leakage points described in the embodiments of the present application refer to the position points on the pipeline to be measured where leakage may exist. Specifically, it can be preliminarily judged by using the temperature information along the pipeline to be measured and determined by quickly identifying temperature outliers.

[0028] The multi-source monitoring information described in the embodiments of the present application refers to the information collected by monitoring signals of suspected leakage points on the pipeline to be measured by using a variety of distributed optical fiber sensing technologies. Specifically, it can include temperature information, vibration information, and strain information.

[0029] It can be understood that the time-frequency characteristic information includes time-domain characteristic information and frequency-domain characteristic information; the time-frequency characteristic information of the multi-source monitoring information is the characteristic information obtained by extracting the time-domain characteristics or frequency-domain characteristics of the temperature information, vibration information, and strain information.

[0030] The leakage detection model described in the embodiments of the present application is trained by using the time-frequency characteristic information of the multi-source monitoring information samples and the corresponding pipeline leakage status labels for a machine learning classification model. Among them, the machine learning classification model can adopt a random forest model, a decision tree model, a linear regression model, etc.

[0031] It can be understood that the multi-source monitoring information samples can specifically include temperature information samples, vibration information samples, and strain information samples.

[0032] In the embodiments of the present application, the pipeline leakage status labels are pre-determined according to the multi-source monitoring information samples and are in one-to-one correspondence with the multi-source monitoring information samples. That is to say, for each multi-source monitoring information sample in the training samples, a corresponding pipeline leakage status label is preset.

[0033] Among them, the pipeline leakage status labels can include five types of labels: the normal operation state of the water supply pipeline when there is no leakage, the leakage state of the water supply pipeline caused by corrosion, the leakage state of the water supply pipeline caused by rupture, the leakage state of the water supply pipeline caused by pipeline distortion, and the leakage state of the water supply pipeline caused by cavitation.

[0034] In the embodiments of the present application, the distributed fiber optic temperature sensing (DTS) technology is used to monitor the temperature changes around the water supply pipeline to be measured, the distributed fiber optic vibration sensing (DAS) technology is used to monitor the vibration signals on the water supply pipeline to be measured, and the distributed fiber optic strain sensing (DSS) technology is used to monitor the strain signals on the water supply pipeline to be measured.

[0035] Specifically, in the embodiments of the present application, the data acquisition method includes: laying multiple types of distributed fiber optic sensors along the water supply pipeline to be measured to ensure full coverage of the pipeline area to be detected.

[0036] When using the DTS system to collect temperature information, the pulsed laser emits optical signals at a certain frequency, which are transmitted to the distributed fiber optic sensor through the wavelength division multiplexer. The sensor exchanges heat with the pipeline environment, and the reflected light is separated by the wavelength division multiplexer. The anti-Stokes light carries temperature information, and the photodetector converts and amplifies the electrical signal, which is collected and transmitted to the computer for processing through the high-speed data acquisition card.

[0037] When using the DAS system to collect vibration information, the phase-sensitive optical time domain reflectometer (Φ - OTDR) technology is used. The pulsed laser emits light pulses that are transmitted in the optical fiber. When encountering vibrations, the refractive index of the optical fiber changes, and the phase of the backscattered Rayleigh light changes. The phase change is detected through the heterodyne detection structure to obtain the vibration signal, which is collected and transmitted to the computer for processing through the data acquisition card.

[0038] When using the DSS system to collect strain information, based on the Brillouin optical time domain analysis (BOTDA) technology, the pulsed laser emits optical pulses, which enter the optical fiber through the wavelength division multiplexer and interact with the pipeline strain to generate Brillouin frequency shift. The backscattered Brillouin light returns, is converted into an electrical signal by the photodetector, and the strain signal is obtained, which is collected and transmitted to the computer for processing in real time through the data acquisition card.

[0039] Further, in the embodiments of the present application, in step S1, through the above method of using the DTS system to collect temperature information, the temperature information along the pipeline to be measured can be obtained, and then by identifying the temperature abnormal values in the temperature information along the pipeline to be measured, the suspected leakage points existing along the pipeline to be measured can be determined. Here, the suspected leakage points can be a certain location or multiple locations along the pipeline to be measured.

[0040] In an embodiment of the present application, in step S2, after determining the suspected leakage points existing along the pipeline to be measured, the above DTS system, DAS system, and DSS system can be further used to centrally monitor the information of the suspected leakage points, obtain multi-source monitoring information of the suspected leakage points, including temperature information, vibration information, and strain information. Furthermore, through time-domain feature calculation and time-frequency feature calculation, the corresponding signal features of various types of monitoring information are extracted, and thus the time-frequency feature information of the multi-source monitoring information is obtained. In an embodiment of the present application, before performing step S3, it is necessary to use the time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage state labels for model training. Using the time-frequency feature data of the multi-source monitoring information under different pipeline leakage states, the machine learning model is made to learn the internal relationship between the time-frequency features of the monitoring information and the leakage state of the water supply pipeline through iterative training, so as to improve the recognition accuracy of the model for the leakage state types of the water supply pipeline. Finally, the trained model is used as the pipeline leakage detection model.

[0041] Furthermore, in step S3, the time-frequency feature information of the multi-source monitoring information obtained in step S2 is input into the pipeline leakage detection model. Through the recognition of the leakage state type of the water supply pipeline by the pipeline leakage detection model, the leakage state information of the suspected leakage points can be obtained, so as to accurately identify whether there is leakage at the suspected leakage points and the leakage state type corresponding to the leakage. The whole process is efficient and highly reliable.

[0042] The water supply pipeline leakage detection method of the embodiment of the present application, by integrating the advantages of various monitoring signals, considering the internal influence relationship between pipeline leakage and each monitoring signal, after locking the suspected leakage points along the pipeline to be measured, uses multi-source monitoring information including temperature information, vibration information, and strain information, and fuses various time-frequency features of the multi-source monitoring information to more comprehensively detect and identify the leakage state of the suspected leakage points. Without a large amount of manual participation, it can effectively improve the pipeline leakage detection accuracy and reliability, while improving the detection efficiency and reducing the labor cost.

[0043] Based on the content of the above embodiment, as an alternative embodiment, based on the multi-source monitoring information of the suspected leakage points, determining the time-frequency feature information of the multi-source monitoring information includes: According to the temperature information of the suspected leakage point, determining various types of time-domain feature information corresponding to the temperature information; According to the vibration information of the suspected leakage point, determining various types of time-frequency feature information corresponding to the vibration information; According to the strain information of the suspected leakage point, determining various types of time-domain feature information corresponding to the strain information; Determine the time-frequency feature information of multi-source monitoring information based on multiple types of time-domain feature information corresponding to temperature information, multiple types of time-frequency feature information corresponding to vibration information, and multiple types of time-domain feature information corresponding to strain information.

[0044] Specifically, in the embodiments of the present application, the implementation manner of determining the time-frequency feature information of multi-source monitoring information is as follows: Using the existing video feature calculation formula, according to the temperature information of the suspected leakage point, multiple types of time-domain feature information corresponding to the temperature information can be calculated; according to the vibration information of the suspected leakage point, multiple types of time-frequency feature information corresponding to the vibration information can be calculated; according to the strain information of the suspected leakage point, multiple types of time-domain feature information corresponding to the strain information can be calculated.

[0045] Based on the content of the above embodiments, as an optional embodiment, multiple types of time-domain feature information corresponding to temperature information include maximum value (MAX), average value (AVE), time-domain range (TR), standard deviation (TSD), third-order origin moment (3OD), third-order central moment (3CD), skewness (TSK), kurtosis (TK); Multiple types of time-frequency feature information corresponding to vibration information include time-domain range (TR), standard deviation (TSD), spectral standard deviation (SSD), spectral skewness (SS), spectral kurtosis (SK), spectral entropy (SE), strongest frequency (SF), ratio of main peak to secondary peak (RMS); Multiple types of time-domain feature information corresponding to strain information include maximum value (MAX), minimum value (MIN), average value (AVE), time-domain range (TR), standard deviation (TSD), root mean square (RMS), waveform index (Peak / RMS), skewness (Skewness), kurtosis (Kurtosis), pulse index (Peak / AVE), zero-crossing rate (Zero-Crossing Rate).

[0046] Specifically, in the embodiments of the present application, the temperature information of the suspected leakage point collected within a specific time period (such as 20 minutes, 30 minutes, etc.) can be processed according to the calculation formula of each time-frequency feature, such as calculating the maximum value, average value, time-domain range, standard deviation, third-order origin moment, third-order central moment, skewness, and kurtosis of the temperature information. For example, when calculating the standard deviation, first find the sum of the squares of the differences between the sampling point temperatures and the average value, and then take the square root after dividing by the number of sampling points minus 1; other time-domain features, such as the third-order origin moment, third-order central moment, skewness, and kurtosis, can be calculated according to the existing calculation formulas.

[0047] Here, for the time-domain feature values extracted from the temperature information, including the maximum value, average value, time-domain range, standard deviation, third-order origin moment, third-order central moment, skewness, and kurtosis, these feature values can intuitively reflect the change of the pipeline temperature.

[0048] In the embodiments of the present application, for the extraction of time-frequency eigenvalues of vibration information, the time-domain features include the time-domain range and the standard deviation, and their calculation methods are the same as those of the time-domain range and the standard deviation of the above temperature information; for the frequency-domain features, such as the spectral standard deviation, skewness, kurtosis, and spectral entropy, the vibration signal needs to be subjected to FFT transformation to obtain the corresponding spectral data first, and then calculated according to the power distribution and the calculation formulas of each frequency-domain feature; for other frequency-domain features, such as the strongest frequency, it can be determined by finding the maximum power point of the spectrum, and the ratio of the main peak to the secondary peak can be calculated by comparing the amplitudes of the maximum power point and the second maximum power point.

[0049] Here, for the eigenvalues extracted from vibration data, leaks caused by corrosion usually manifest as low-frequency vibrations, with a concentrated spectrum and low complexity, a small time-domain range and standard deviation, a negative spectral skewness, and a prominent main peak; ruptures manifest as rich high-frequency components, a large time-domain range and standard deviation, a positive spectral skewness, scattered energy, and multiple peaks coexisting; pipe distortion is mainly low-frequency periodic vibration, with a sharp spectrum and highly concentrated energy, and a prominent main peak; cavitation manifests as high-frequency shock waves, with a wide spectral distribution and high complexity, and an insignificant main peak.

[0050] In the embodiments of the present application, for the extraction of time-domain features of strain information, including the maximum value, minimum value, average value, time-domain range, standard deviation, root mean square, waveform index, skewness, kurtosis, pulse index, and zero-crossing rate, they can all be calculated according to the corresponding calculation formulas using the strain information within a specific time period. For example, the waveform index can be obtained by calculating the ratio of the maximum strain information to the root mean square; the pulse index can be obtained by calculating the ratio of the maximum strain information to the average value.

[0051] Here, leaks caused by corrosion usually manifest as relatively low eigenvalues such as the maximum value, average value, standard deviation, and root mean square of the strain signal, and the waveform index, skewness, kurtosis, pulse index, and zero-crossing rate are also relatively small, reflecting that the signal fluctuates gently and stably; ruptures manifest as relatively high eigenvalues such as the maximum value, standard deviation, and root mean square, and the waveform index, skewness, kurtosis, pulse index, and zero-crossing rate increase significantly, and the signal fluctuates violently and has obvious mutations; the eigenvalues of pipe distortion are usually between those of corrosion and rupture, and the signal fluctuates relatively smoothly and has strong periodicity; cavitation manifests as relatively high eigenvalues such as the maximum value, standard deviation, and root mean square, and the waveform index, skewness, kurtosis, pulse index, and zero-crossing rate increase significantly, and the signal fluctuates complexly and has high-frequency impact characteristics. These characteristic differences can be used to effectively distinguish different diseases causing pipeline leaks.

[0052] The method according to the embodiments of the present application can, by deeply mining the relationship between the time-domain features or frequency-domain features of various monitoring information and the pipeline leakage state, make full use of the features that can reflect the changes in parameters such as temperature, vibration, and strain during pipeline leakage for comprehensive analysis, which helps to identify potential leakage points of the pipeline and improve the accuracy and reliability of subsequent pipeline leakage monitoring results.

[0053] Furthermore, in the embodiments of the present application, based on the multiple types of time-domain feature information corresponding to the above-extracted temperature information, the multiple types of time-frequency feature information corresponding to the vibration information, and the multiple types of time-domain feature information corresponding to the strain information, feature information fusion is performed. All the above features can be spliced into a fused feature vector, thereby obtaining the time-frequency feature information of multi-source monitoring information.

[0054] The method according to the embodiments of the present application can, by extracting multiple types of time-domain feature information or frequency-domain feature information corresponding to various monitoring signals for the fusion of multi-source time-frequency feature information, capture the subtle changes in the monitoring signals in terms of time and frequency. These changes are often closely related to leakage events, which is conducive to further improving the accuracy and reliability of subsequent pipeline leakage detection.

[0055] Based on the content of the above embodiments, as an optional embodiment, after determining the multiple types of time-domain feature information corresponding to the strain information according to the strain information of the suspected leakage point, the method further includes: Determining the feature information of multiple types of first target features corresponding to the temperature information from the multiple types of time-domain feature information corresponding to the temperature information; Determining the feature information of multiple types of second target features corresponding to the vibration information from the multiple types of time-frequency feature information corresponding to the vibration information; Determining the feature information of multiple types of third target features corresponding to the strain information from the multiple types of time-domain feature information corresponding to the strain information; the multiple types of first target features, the multiple types of second target features, and the multiple types of third target features are determined by performing time-frequency feature screening based on multi-source monitoring information samples and corresponding pipeline leakage state labels; Based on the feature information of each first target feature, the feature information of each second target feature, and the feature information of each third target feature, determining the time-frequency feature information of multi-source monitoring information.

[0056] Specifically, the first target feature described in the embodiments of the present application refers to the time-domain features with significant representativeness screened from multiple types of time-domain features corresponding to the temperature information by using statistical analysis means of feature data.

[0057] The second target feature described in the embodiments of the present application refers to the time-domain features and / or frequency-domain features with significant representativeness screened from multiple types of time-frequency features corresponding to the vibration information by using statistical analysis means of feature data.

[0058] The third target features described in the embodiments of the present application respectively refer to the time-domain features with significant representativeness selected from multiple types of time-domain features corresponding to strain information by using statistical analysis means of feature data.

[0059] In the embodiments of the present application, statistical analysis means can be adopted in advance to perform time-frequency feature screening by using multi-source monitoring information samples and corresponding pipeline leakage status labels, and select multiple types of first target features, multiple types of second target features, and multiple types of third target features from multiple types of time-domain feature information corresponding to temperature information, multiple types of time-frequency feature information corresponding to vibration information, and multiple types of time-domain feature information corresponding to strain information, so as to reduce the amount of calculation data and improve the efficiency of subsequent model detection.

[0060] Based on the content of the above embodiments, as an optional embodiment, before determining the time-frequency feature information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point, the method further includes: Determine multiple types of time-domain feature information of the temperature information samples, multiple types of time-frequency feature information of the vibration information samples, and multiple types of time-domain feature information of the strain information samples corresponding to each pipeline leakage status label according to the multi-source monitoring information samples corresponding to each pipeline leakage status label; For any feature information among the multiple types of time-domain feature information of the temperature information samples, multiple types of time-frequency feature information of the vibration information samples, and multiple types of time-domain feature information of the strain information samples, determine the Minkowski distance between any feature information under different pipeline leakage states; According to the Minkowski distance between any feature information under different pipeline leakage states, determine whether the feature corresponding to any feature information is a fourth target feature, so as to determine multiple types of fourth target features and the remaining features other than the multiple types of fourth target features; Perform rank correlation coefficient analysis on the feature information of the remaining features, and determine multiple types of fifth target features from the remaining features; Determine multiple types of first target features, multiple types of second target features, and multiple types of third target features according to the multiple types of fourth target features and multiple types of fifth target features.

[0061] Specifically, the fourth target features described in the embodiments of the present application refer to the features with high discrimination ability selected from multiple types of time-frequency features of multi-source monitoring information samples by using the Minkowski distance judgment.

[0062] The fifth target features described in the embodiments of the present application refer to the features with high discrimination ability selected from the remaining features by rank correlation coefficient analysis.

[0063] In an embodiment of the present application, according to the multi-source monitoring information samples corresponding to each pipeline leakage status label, the multi-class time-domain feature information of the temperature information samples, the multi-class time-frequency feature information of the vibration information samples, and the multi-class time-domain feature information of the strain information samples corresponding to each pipeline leakage status label are calculated according to the formula of the aforementioned time-frequency characteristics.

[0064] Further, for any one of the multi-class time-domain feature information of the temperature information samples, the multi-class time-frequency feature information of the vibration information samples, and the multi-class time-domain feature information of the strain information samples, by adopting the Minkowski distance feature screening method, calculate the Minkowski distance of the eigenvalues of each type of feature under different water supply pipeline states, classify the data, calculate the distance according to the formula, compare the magnitudes of the distances to judge the discrimination ability of the eigenvalues, and delete the eigenvalues with poor discrimination ability.

[0065] The specific calculation process is as follows: For each feature, regard the feature information under different pipeline states as different event categories, and assume two columns of feature vectors and are respectively composed of multiple values of the same feature in two different event categories and , and their components can be expressed as and , represents the serial number of the eigenvalue in the vector, and the formula is as follows: , (where = 3); In the formula, represents the Minkowski distance of the same feature under two different event categories and .

[0066] For example, for the time-domain feature of the maximum value in the temperature data, if a set of data in the normal operation state of the water supply pipeline when the label is non-leakage is (20, 21, 22), a set of data in the water supply pipeline leakage state caused by corrosion when the label is corrosion is (26, 28, 27), and a set of data in the water supply pipeline leakage state caused by cavitation when the label is cavitation is (24, 26, 25), then the Minkowski distance between any two of the different state labels of this time-domain feature can be calculated. By comparing the magnitude differences among the three calculated Minkowski distances, if there are large differences, it can be determined that the time-domain feature of the maximum value of the temperature data has high discrimination ability for the three states of the normal pipeline state, corrosion leakage state, and cavitation leakage state. Thus, this time-domain feature can be used as the fourth target feature.

[0067] Thus, in the above manner, according to the Minkowski distance between any characteristic information under different pipeline leakage states, it can be determined whether the characteristic is the fourth target characteristic, and then multiple types of fourth target characteristics can be determined, as well as the remaining characteristics other than multiple types of fourth target characteristics.

[0068] Furthermore, using the Spearman rank correlation coefficient, rank correlation coefficient analysis is performed based on the characteristic information of the remaining characteristics. The specific calculation process is as follows: Let x and y be column vectors composed of multiple eigenvalues of two different characteristics respectively. and are the averages of elements x and y . The Spearman rank correlation coefficient is determined through the following formula , that is: ; Here, if the calculation result is greater than 0.7, it is considered that the correlation between the two characteristics is high, and any one of the characteristic values can be retained.

[0069] For example, for the two time-domain characteristics of the average value and the third-order origin moment of temperature information data, if the calculated SRCC is greater than 0.7, one of the characteristics can be retained according to the actual situation, and the other can be eliminated. In this way, multiple types of fifth target characteristics can be determined from the remaining characteristics, thus achieving the effect of simplifying the feature vector data.

[0070] Finally, according to the multiple types of fourth target characteristics and multiple types of fifth target characteristics obtained above, multiple types of first target characteristics corresponding to temperature information, multiple types of second target characteristics corresponding to vibration information, and multiple types of third target characteristics corresponding to strain information can be directly obtained, thereby obtaining an optimized time-frequency feature combination.

[0071] The method of the embodiment of the present application, by combining the Minkowski distance and rank correlation coefficient analysis, uses the Minkowski distance of time-frequency characteristics under different pipeline leakage states to screen target characteristics, and then calculates the Spearman rank correlation coefficient between the remaining characteristic values, further optimizing the time-frequency feature combination of multi-source monitoring information, which is beneficial to reducing the amount of characteristic calculation data of the model.

[0072] Furthermore, in the embodiments of the present application, in the actual application scenario, from the multiple types of time-domain feature information corresponding to the temperature information, the feature information of multiple types of first target features corresponding to the temperature information can be determined; from the multiple types of time-frequency feature information corresponding to the vibration information, the feature information of multiple types of second target features corresponding to the vibration information can be determined; and from the multiple types of time-domain feature information corresponding to the strain information, the feature information of multiple types of third target features corresponding to the strain information can be determined. Finally, the feature information of each first target feature, the feature information of each second target feature, and the feature information of each third target feature can be combined to obtain the time-frequency feature information of the multi-source monitoring information.

[0073] The method of the embodiments of the present application optimizes and simplifies the time-frequency features of the multi-source monitoring information, and selects several representative target features from the time-frequency features of multiple types of multi-source monitoring information for detection, which can greatly reduce the data volume of feature calculation and improve the calculation efficiency of the pipeline leakage detection model.

[0074] Based on the content of the above embodiments, as an alternative embodiment, before inputting the time-frequency feature information of the multi-source monitoring information into the pipeline leakage detection model to obtain the leakage state information of the suspected leakage point output by the pipeline leakage detection model, the method further includes: Taking the time-frequency feature information of the multi-source monitoring information sample and the corresponding pipeline leakage state label as a set of training samples, and obtaining multiple sets of training samples; Training the random forest model with multiple sets of training samples to obtain a trained random forest model, and using the trained random forest model as the pipeline leakage detection model.

[0075] Specifically, in the embodiments of the present application, a random forest model is used to detect and identify the leakage of the water supply pipeline. Before inputting the time-frequency feature information of the multi-source monitoring information into the pipeline leakage detection model to obtain the leakage state information of the suspected leakage point output by the pipeline leakage detection model, the random forest model also needs to be trained to obtain a trained random forest model and obtain the pipeline leakage detection model.

[0076] In the embodiments of the present application, the temperature data, vibration data, and strain data under different pipeline leakage states are collected as multi-source monitoring information samples. Then, according to the aforementioned feature extraction and screening methods, the time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage state labels can be obtained, so that the obtained feature vector data can be composed into a training data set.

[0077] Furthermore, a random forest model is used to construct a pipeline leakage detection model, parameters such as the number of decision trees are set, and a 10-fold cross-validation method is adopted. The training data set is divided into ten parts, tested and trained alternately, and the average accuracy can be used to evaluate the model, and the training and testing time are recorded.

[0078] More specifically, in an embodiment of the present invention, the random forest model is trained using the above training dataset, and the specific training process is as follows: The time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage status labels are used as a set of training samples, that is, the time-frequency feature information of each multi-source monitoring information sample with a pipeline leakage status label is used as a set of training samples, and thus multiple sets of training samples can be obtained.

[0079] In an embodiment of the present application, the multi-source monitoring information samples and the pipeline leakage status labels they carry are in one-to-one correspondence.

[0080] Then, after obtaining multiple sets of training samples, the multiple sets of training samples are sequentially input into the random forest model, and the random forest model is trained using the multiple sets of training samples, that is: Using the random sampling method with replacement, multiple subsets are drawn from the multiple sets of training samples, and each subset is used to train a decision tree. During the sampling process, samples are allowed to repeat to ensure that the training samples of each decision tree are independent and diverse. For each drawn sample subset, a decision tree is constructed using the fully split method. Furthermore, when splitting each node of the decision tree, the optimal feature is selected from the randomly selected features for splitting. The growth process of the decision tree continues until the stop condition is met, for example, reaching the maximum depth, the number of samples in the node is less than a predetermined threshold, etc. All the trained decision trees are combined into a trained random forest model, that is, a pipeline leakage detection model is obtained. When new input data is input into the pipeline leakage detection model, each decision tree will perform classification or regression prediction on it.

[0081] The method of the embodiment of the present application, by using the time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage status labels as a set of training samples, and training the random forest model using multiple sets of training samples, is beneficial to improving the model accuracy of the trained random forest model.

[0082] Figure 2 It is the second flowchart of the water supply pipeline leakage detection method provided by the embodiment of the present application, as Figure 2As shown in the figure, in the embodiment of the present application, first, the DAS system is used to preliminarily locate and determine the suspected leakage points along the pipeline by judging the temperature difference. Then, the DTS system and the DSS system are used to extract the multi-source monitoring information corresponding to the suspected leakage points, including temperature, vibration and strain information. Further, the time-domain features of the temperature information, the time-frequency features of the vibration information, and the time-domain features of the strain information are extracted. Furthermore, from the above features, according to the multi-class target features pre-screened by analyzing the Minkowski distance and the rank correlation coefficient, the eigenvalue of each target feature is determined, and thus a fusion feature vector is formed and input into the pipeline leakage detection model to judge the leakage situation of the suspected leakage points, so as to accurately judge whether there is leakage at the suspected leakage points, and when there is leakage, the corresponding leakage state type, including the leakage state of the water supply pipeline caused by corrosion, the leakage state of the water supply pipeline caused by rupture, the leakage state of the water supply pipeline caused by pipeline distortion, and the leakage state of the water supply pipeline caused by cavitation. Finally, the pipeline condition assessment and repair suggestions can be further carried out on the pipe segments with leakage.

[0083] In a specific embodiment of the present application, after determining the existence of leakage at the suspected leakage point and its corresponding leakage state type, the disease type of the leakage at the pipeline leakage point can be known, and the disease damage degree of the leakage point can be further judged.

[0084] Table 1

[0085] Figure 3 is a schematic flow chart of the assessment of the leakage condition of the water supply pipeline provided by the embodiment of the present application. As Figure 3 shown, in the embodiment of the present application, after determining the existence of leakage at the suspected leakage point, the calculation of the pipe section repair index PRI can be carried out to evaluate the pipeline disease condition of the pipe section where the leakage point is located and put forward repair suggestions.

[0086] Specifically, first, the pipe section of the leakage point itself and its surrounding environment are checked, and then the diseases are scored according to Table 1 above: Next, the number of pipe section diseases n is counted. On the one hand, the pipeline damage parameters S and Smax can be calculated according to the following calculation method, and then the pipeline defect parameter F can be determined, and the pipeline disease level can be determined according to Table 2.

[0087] Table 2

[0088] Among them, the pipeline disease parameters are calculated according to the following formula: When Smax≥S, F = Smax; when Smax<S, F = S; In the formula: F represents the pipeline disease parameter; Smax represents the pipe section damage condition parameter, which is the score at the most severely damaged part of the pipe section disease; S represents the pipe section damage condition parameter, which is the average score calculated according to the number of disease points.

[0089] Among them, the pipe section damage condition parameter S is calculated according to the following formula: ; ; ; In the formula: n represents the number of diseases in the pipe section; n1 represents the number of diseases with a longitudinal net distance greater than 1.5m; n2 represents the number of diseases with a longitudinal net distance greater than 1.0m and not greater than 1.5m. represents the disease score with a longitudinal net distance greater than 1.5m, and takes values according to Table 1; represents the disease score with a longitudinal net distance greater than 1.0m and not greater than 1.5m, and takes values according to Table 1; α represents the disease influence coefficient, which is related to the disease spacing. When the longitudinal net distance of the disease is greater than 1.0m and not greater than 1.5m, α = 1.1.

[0090] Table 3

[0091] On the other hand, determine the pipeline length L and the disease length, and calculate the disease density . Among them, the disease density can be calculated according to the following formula: ; In the formula: represents the disease density; L represents the pipe section length (m); represents the disease length (m) with a longitudinal net distance greater than 1.5m; represents the disease length (m) with a longitudinal net distance greater than 1.0m and not greater than 1.5m.

[0092] Furthermore, determine the pipeline disease degree according to Table 3 above.

[0093] Table 4

[0094] Table 5

[0095] Among them, after determining the pipeline defect parameter F, the pipe section repair index PRI can also be calculated according to the following calculation formula, and the maintenance level and suggestions can be further determined according to Table 7 below: P ; Wherein, PRI represents the pipe section repair index; K represents the regional importance parameter, which can be determined with reference to Table 4 above; E represents the pipeline importance parameter, which can be determined with reference to Table 5 above; T represents the soil quality influence parameter, which can be determined with reference to Table 6 below.

[0096] Table 6

[0097] Table 7

[0098] The water supply pipeline leakage detection device provided by the present application will be described below. The water supply pipeline leakage detection device described below can be correspondingly referred to the water supply pipeline leakage detection method described above.

[0099] Figure 4 is a schematic structural diagram of the water supply pipeline leakage detection device provided by an embodiment of the present application. As Figure 4 shown, it includes: The first processing module 10 is used to determine the suspected leakage points along the pipeline to be measured based on the temperature information along the pipeline to be measured; The second processing module 20 is used to determine the time-frequency characteristic information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage points; The leakage detection module 30 is used to input the time-frequency characteristic information of the multi-source monitoring information into the pipeline leakage detection model to obtain the leakage state information of the suspected leakage points output by the pipeline leakage detection model; The multi-source monitoring information includes temperature information, vibration information, and strain information; the leakage detection model is trained according to the time-frequency characteristic information of the multi-source monitoring information samples and the corresponding pipeline leakage state labels.

[0100] It can be understood that the detailed function implementation of each of the above units / modules can be referred to the introduction in the foregoing method embodiments, and will not be elaborated here.

[0101] It should be understood that the above device is used to execute the method in the above embodiment. For the corresponding program modules in the device, their implementation principles and technical effects are similar to those described in the above method. The working process of this device can refer to the corresponding process in the above method, and will not be elaborated here.

[0102] The water supply pipeline leakage detection device according to the embodiment of the present application comprehensively utilizes the advantages of various monitoring signals, considers the internal influence relationship between pipeline leakage and each monitoring signal, and after locking the suspected leakage points along the pipeline to be measured, uses multi-source monitoring information including temperature information, vibration information and strain information, and fuses various time-frequency characteristics of the multi-source monitoring information to more comprehensively detect and identify the leakage state of the suspected leakage points. Without a large amount of manual participation, it can effectively improve the accuracy and reliability of pipeline leakage detection, while improving the detection efficiency and reducing the labor cost.

[0103] Based on the method in the above embodiment, the embodiment of the present application provides an electronic device, as Figure 5 shown. The electronic device may include: a processor 510, a communication interface 520, a memory 530, and a communication bus 540. Among them, the processor 510, the communication interface 520, and the memory 530 communicate with each other through the communication bus 540. The processor 510 can call the logical instructions in the memory 530 to execute the method in the above embodiment.

[0104] In addition, when the logical instructions in the above-mentioned memory 530 are implemented in the form of software function units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application.

[0105] Based on the method in the above embodiment, the embodiment of the present application provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program runs on a processor, it causes the processor to execute the method in the above embodiment.

[0106] Based on the method in the above embodiment, the embodiment of the present application provides a computer program product. When the computer program product runs on a processor, it causes the processor to execute the method in the above embodiment.

[0107] It can be understood that the processor in the embodiments of the present application may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0108] The method steps in the embodiments of the present application may be implemented in a hardware manner or by a processor executing software instructions. The software instructions may be composed of corresponding software modules, and the software modules may be stored in a random access memory (RAM), flash memory, read-only memory (ROM), programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), registers, hard disks, removable hard disks, CD-ROMs, or any other form of storage medium well known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium may also be a component of the processor. The processor and the storage medium may be located in the ASIC.

[0109] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)), etc.

[0110] It can be understood that the various numerical numbers involved in the embodiments of the present application are only for the convenience of description and are not used to limit the scope of the embodiments of the present application.

[0111] It should be understood that expressions such as "including" and "may include" that can be used in the present application indicate the existence of the disclosed functions, operations, or constituent elements, and do not limit the existence of one or more additional functions, operations, and constituent elements. In the present application, terms such as "including" and / or "having" can be interpreted as indicating a specific characteristic, number, operation, constituent element, component, or a combination thereof, but cannot be interpreted as excluding the existence or possibility of addition of one or more other characteristics, numbers, operations, constituent elements, components, or a combination thereof.

[0112] As described above, the above are only the specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A water supply pipeline leakage detection method, characterized in that: include: Determine suspected leakage points along the pipeline to be tested based on temperature information along the pipeline to be tested; Based on the multi-source monitoring information of the suspected leakage point, determining the time-frequency characteristic information of the multi-source monitoring information; Inputting the time-frequency characteristic information of the multi-source monitoring information into a pipeline leakage detection model to obtain leakage status information of the suspected leakage point output by the pipeline leakage detection model; The multi-source monitoring information includes temperature information, vibration information and strain information; the leakage detection model is trained based on the time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage state labels.

2. The water supply pipeline leakage detection method according to claim 1, characterized in that: The determining the time-frequency characteristic information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point includes: Determine, according to the temperature information of the suspected leakage point, multiple types of time domain feature information corresponding to the temperature information; Determine, according to the vibration information of the suspected leakage point, multiple types of time-frequency feature information corresponding to the vibration information; Determining, according to the strain information of the suspected leakage point, multiple types of time domain feature information corresponding to the strain information; The time-frequency feature information of the multi-source monitoring information is determined based on the multiple types of time-domain feature information corresponding to the temperature information, the multiple types of time-frequency feature information corresponding to the vibration information, and the multiple types of time-domain feature information corresponding to the strain information.

3. The water supply pipeline leakage detection method according to claim 2, characterized in that: After determining multiple types of time domain feature information corresponding to the strain information according to the strain information of the suspected leakage point, the method further includes: Determining, from the multiple types of time-domain feature information corresponding to the temperature information, feature information of multiple types of first target features corresponding to the temperature information; Determining, from the multiple types of time-frequency feature information corresponding to the vibration information, feature information of multiple types of second target features corresponding to the vibration information; Determine, from the multiple categories of time-domain feature information corresponding to the strain information, feature information of multiple categories of third target features corresponding to the strain information; the multiple categories of the first target features, the multiple categories of the second target features, and the multiple categories of the third target features are determined by screening the time-frequency features based on the multi-source monitoring information samples and the corresponding pipeline leakage status labels; Based on the feature information of each of the first target features, the feature information of each of the second target features, and the feature information of each of the third target features, the time-frequency feature information of the multi-source monitoring information is determined.

4. The water supply pipeline leakage detection method according to claim 3, characterized in that: Before determining the time-frequency feature information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point, the method further includes: According to the multi-source monitoring information samples corresponding to each pipeline leakage status label, determine the multi-class time domain feature information of the temperature information sample corresponding to each pipeline leakage status label, the multi-class time-frequency feature information of the vibration information sample, and the multi-class time domain feature information of the strain information sample; For any feature information among the multiple types of time-domain feature information of the temperature information sample, the multiple types of time-frequency feature information of the vibration information sample, and the multiple types of time-domain feature information of the strain information sample, determining the Minkowski distance between any feature information under different pipeline leakage states; Determine whether a feature corresponding to any feature information is a fourth target feature according to the Minkowski distance between any feature information under different pipeline leakage states, so as to determine multiple types of the fourth target features and remaining features other than the multiple types of the fourth target features; Performing rank correlation coefficient analysis based on the feature information of the remaining features, and determining multiple categories of fifth target features from the remaining features; According to the multiple categories of the fourth target features and the multiple categories of the fifth target features, multiple categories of the first target features, multiple categories of the second target features and multiple categories of the third target features are determined.

5. The water supply pipeline leakage detection method according to claim 2, characterized in that: The multiple types of time domain feature information corresponding to the temperature information include maximum value, average value, time domain range, standard deviation, third-order origin moment, third-order central moment, skewness and kurtosis; The multiple types of time-frequency feature information corresponding to the vibration information include time domain range, standard deviation, spectrum standard deviation, spectrum skewness, spectrum kurtosis, spectrum entropy, strongest frequency, and ratio of main peak to secondary peak; The multiple types of time domain feature information corresponding to the strain information include maximum value, minimum value, average value, time domain range, standard deviation, root mean square, waveform index, skewness, kurtosis, pulse index and zero crossing rate.

6. The water supply pipeline leakage detection method according to any one of claims 1 to 5, characterized in that: Before inputting the time-frequency characteristic information of the multi-source monitoring information into the pipeline leakage detection model to obtain the leakage status information of the suspected leakage point output by the pipeline leakage detection model, the method further includes: Taking the time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage status labels as a group of training samples, and obtaining multiple groups of the training samples; The random forest model is trained using multiple groups of the training samples to obtain a trained random forest model, and the trained random forest model is used as the pipeline leakage detection model.

7. A water supply pipeline leakage detection device, characterized in that: include: A first processing module, configured to determine a suspected leakage point along the pipeline to be tested based on temperature information along the pipeline to be tested; A second processing module, configured to determine time-frequency feature information of the multi-source monitoring information based on the multi-source monitoring information of the suspected leakage point; A leakage detection module, used to input the time-frequency characteristic information of the multi-source monitoring information into a pipeline leakage detection model, and obtain leakage status information of the suspected leakage point output by the pipeline leakage detection model; The multi-source monitoring information includes temperature information, vibration information and strain information; the leakage detection model is trained based on the time-frequency feature information of the multi-source monitoring information samples and the corresponding pipeline leakage state labels.

8. An electronic device, characterized in that: include: at least one memory for storing a computer program; At least one processor is used to execute the program stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program runs on a processor, the processor is caused to execute the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that When the computer program product runs on a processor, the processor is caused to execute the method according to any one of claims 1 to 6.

Images