Water supply network leakage monitoring and early warning method and system based on water power noise, medium and equipment

Abstract

The present application relates to the field of city water supply pipe network leakage early warning, discloses a kind of water supply pipe network leakage monitoring early warning method, system, medium and equipment based on water power noise, it includes: determining monitoring site, analysis water power noise source and generation site, to simulate the generation of water power noise;Water power noise data of monitoring site is obtained, water power noise data is filtered noise processing, and feature extraction is carried out, and the time-frequency domain feature extracted is used as the representation parameter of acoustic signal;Based on the time-frequency domain feature extracted, a leakage identification early warning model is constructed using a random forest, to determine whether the fluid flow in the pipe is abnormal, to identify and warn the abnormal state.The present application analyzes the acoustic vibration caused by the leakage of water supply pipe network in depth, identifies and warns the leakage based on the characteristics of water power noise, and effectively monitors the leakage events within the listening coverage of the sensor.

Description

Technical Field

[0001] This invention relates to the field of leakage early warning technology for urban water supply networks, and in particular to a method, system, medium, and equipment for monitoring and early warning of leakage in water supply networks based on hydrodynamic noise. Background Technology

[0002] Leakage detection and early warning technology for water supply networks is a crucial component of modern urban infrastructure maintenance and management. With the continuous advancement of urbanization, water supply networks are becoming increasingly large and complex, and pipeline leaks are becoming more prominent. Pipeline leaks not only waste precious water resources but can also damage underground infrastructure and cause environmental problems. Therefore, effectively identifying leak points in water supply networks has become an important aspect of urban water supply system management and maintenance. The large scale, complex structure, and diverse pipe materials of modern urban water supply networks increase the difficulty of leak detection. Traditional manual inspection and pressure monitoring methods are no longer sufficient to meet the demands for efficient and accurate detection. Frequent leaks in aging pipes not only increase maintenance costs but can also disrupt urban traffic and residents' lives, and even trigger secondary disasters. Therefore, diverse leak detection technologies are needed to address the problem of leak monitoring in complex water supply networks.

[0003] Leakage identification and early warning technologies for water supply networks have evolved and iterated through continuous theoretical research and engineering practice. Existing leakage monitoring and early warning technologies for water supply networks include, but are not limited to, pressure data monitoring, flow data monitoring, pipeline acoustic monitoring, and pipeline endoscopy. Monitoring and early warning of leakage events in water supply networks requires not only high-precision detection technology but also consideration of the actual operating conditions of the water supply network. The pipeline external wall detection method within acoustic monitoring is a non-destructive testing technique that does not affect the normal operation of the water supply network or the structure of the water supply pipeline system. Utilizing high-precision vibration sensors to collect acoustic signals from water supply pipelines and analyzing relevant characteristics for leakage identification and early warning is a key research area in acoustic monitoring of water supply network leaks.

[0004] Due to the complexity and diversity of water supply pipeline systems, and issues such as noise interference around monitoring wells, acoustic monitoring technology on the outer wall of pipelines faces challenges such as severe interference signals and unclear acoustic signal characteristics. Therefore, a clear understanding of the acoustic vibrations caused by pipeline leaks is needed. This can be achieved through the rational placement of sensors, multiple analyses of acoustic signal characteristics, and a combination of mechanistic understanding and data-driven approaches to realize effective monitoring and early warning of water supply pipeline leaks. Summary of the Invention

[0005] To address the aforementioned problems, the present invention aims to provide a method, system, medium, and equipment for monitoring and early warning of leaks in water supply networks based on hydrodynamic noise. By deeply analyzing the acoustic vibrations caused by leaks in water supply networks, leak identification and early warning are performed based on the characteristics of hydrodynamic noise, thereby achieving effective monitoring of leak events within the sensor's monitoring coverage area.

[0006] To achieve the above objectives, in a first aspect, the technical solution adopted by the present invention is as follows: a method for monitoring and early warning of water supply network leakage based on hydrodynamic noise, comprising: determining monitoring sites, analyzing the sources and generation sites of hydrodynamic noise to simulate the generation of hydrodynamic noise; acquiring hydrodynamic noise data of monitoring sites, performing noise filtering on the hydrodynamic noise data, and extracting features, using the extracted time-frequency domain features as the characterization parameters of the sound signal; constructing a leakage identification and early warning model based on the extracted time-frequency domain features using a random forest, determining whether the fluid flow in the pipe is abnormal, so as to identify and warn of abnormal states.

[0007] Furthermore, based on the upstream and downstream relationship of water flow in the pipeline network, the monitoring sites are selected from the upstream hydrodynamic noise excitation points, and the locations are selected near the hydrodynamic noise source points of elbows and / or tees.

[0008] Furthermore, hydrodynamic noise data from the monitoring sites is acquired, and the hydrodynamic noise data is processed by noise filtering, including:

[0009] Acoustic monitoring sensors are deployed at monitoring sites near pipe fittings to collect hydrodynamic noise data under different conditions, focusing on the low-frequency band.

[0010] Noise filtering is performed on hydrodynamic noise data under different scenarios to remove pipeline interference noise that does not meet the characteristics of hydrodynamic noise, and a hydrodynamic noise dataset is constructed.

[0011] Furthermore, wavelet transform was used to perform wavelet noise filtering on the collected hydrodynamic noise data.

[0012] Furthermore, the time-frequency domain features include: time-domain energy, time-domain average energy, time-domain maximum energy, time-domain kurtosis, time-domain skewness, frequency-domain shape factor, frequency-domain impulse factor, frequency-domain kurtosis factor, and Mel-frequency cepstral coefficients.

[0013] Furthermore, a leak identification and early warning model is constructed using random forest based on the extracted time-frequency domain features, including:

[0014] The extracted time-frequency domain features are compressed and reconstructed to obtain a feature information matrix;

[0015] A leakage identification and early warning model is constructed based on the feature information matrix. Multiple independent identification decision trees are established by adopting a mode of random data combination and random feature extraction.

[0016] The classification of data is determined by combining the identification results of multiple decision trees, and it is then determined whether a leakage event has occurred at the identified monitoring location.

[0017] Secondly, the technical solution adopted by the present invention is as follows: a water supply network leakage monitoring and early warning system based on hydrodynamic noise, comprising: a data parsing module, which determines the monitoring site, analyzes the source and generation site of hydrodynamic noise, and simulates the generation of hydrodynamic noise; a feature extraction module, which acquires the hydrodynamic noise data of the monitoring site, performs noise filtering on the hydrodynamic noise data, and performs feature extraction, using the extracted time-frequency domain features as the characterization parameters of the sound signal; and an early warning module, which constructs a leakage identification and early warning model based on the extracted time-frequency domain features using a random forest, determines whether the fluid flow in the pipe is abnormal, and identifies and warns of abnormal states.

[0018] Furthermore, based on the upstream and downstream relationship of water flow in the pipeline network, the monitoring sites are selected from the upstream hydrodynamic noise excitation points, and the locations are selected near the hydrodynamic noise source points of elbows and / or tees.

[0019] Thirdly, the technical solution adopted by the present invention is: a computer-readable storage medium for storing one or more programs, wherein the one or more programs include instructions, which, when executed by a computing device, cause the computing device to perform any of the methods described above.

[0020] Fourthly, the technical solution adopted by the present invention is: a computing device comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing any of the methods described above.

[0021] The present invention has the following advantages due to the adoption of the above technical solutions:

[0022] 1. Because hydrodynamic noise is generated by the inherent components of the pipeline, its characteristic patterns are more easily identifiable than those of general non-stationary signals. When monitoring the acoustic vibration of pipelines, the vibration caused by a leak is usually related to the characteristics of the leak point, including but not limited to the shape, location, and surrounding soil environment of the leak. This results in the acoustic vibration caused by a leak typically exhibiting instability, variable states, and irregular intensity.

[0023] 2. Leakage events will affect the fluid state within the pipe; therefore, changes in hydrodynamic noise can serve as indirect observation data for leakage events. The essence of a leakage event is fluid loss within the pipe; therefore, fluid loss downstream will be replenished from upstream, resulting in a flow pattern within the pipe that deviates from the normal state. Its monitoring distance is longer and its coverage area is larger than that of general acoustic monitoring.

[0024] 3. The hydrodynamic noise distribution frequency band of this invention is more concentrated than that of general noise, making it suitable for feature extraction and characterization under interference environments. Due to the complex environment around pipeline wells, interference noise is diverse and exhibits randomness and non-steady-state characteristics, noise interference removal is an important aspect of acoustic monitoring. Compared to general non-stationary signals, hydrodynamic noise has a more regular sound source vibration mode, giving it an advantage in data separation. Attached Figure Description

[0025] Figure 1 This is a flowchart of the water supply network leakage monitoring and early warning method based on hydrodynamic noise in an embodiment of the present invention;

[0026] Figure 2 This is a diagram showing the fluid velocity and pressure distribution in the elbow structure in this embodiment of the invention;

[0027] Figure 3 This is the power spectrum of the hydraulic noise of the elbow in an embodiment of the present invention;

[0028] Figure 4 This is a diagram showing the layout of the elbow-tee hydrodynamic noise monitoring system in an embodiment of the present invention.

[0029] Figure 5 This is an analytical diagram of the wavelet denoising results in an embodiment of the present invention;

[0030] Figure 6 This is a diagram of the random forest model structure in an embodiment of the present invention. Detailed Implementation

[0031] To address the problems of severe interference signals and unclear acoustic signal characteristics in existing pipeline external acoustic monitoring technologies, this invention provides a method, system, medium, and equipment for monitoring and early warning of leaks in water supply networks based on hydrodynamic noise. First, the source and generation location of hydrodynamic noise are analyzed, demonstrating a correlation between the intensity characteristics of hydrodynamic noise and the fluid flow state within the pipeline. Pipeline leaks cause the fluid within the pipe to exhibit flow patterns inconsistent with normal conditions, thus affecting the hydrodynamic noise. Hydrodynamic noise data collected from locations such as elbows and tees using vibration sensors are used to determine abnormal fluid flow within the pipe. Data processing and feature extraction are performed on the hydrodynamic noise under both normal and abnormal conditions, and a hydrodynamic noise anomaly early warning method is established using model recognition. The hydrodynamic noise monitored by this invention is relatively stable sound data, significantly different from other pipeline interference noises, which is beneficial for data noise reduction and feature extraction. Simultaneously, the monitoring range covers the downstream area of ​​the monitoring point, providing a wide monitoring coverage area.

[0032] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0033] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0034] In one embodiment of the present invention, a method for monitoring and early warning of leaks in water supply networks based on hydrodynamic noise is provided. This method uses a high-sensitivity vibration sensor to monitor hydrodynamic noise and analyzes the characteristics of the collected data to provide early warning of leaks in the water supply network. In this embodiment, as... Figure 1 As shown, the method includes the following steps:

[0035] 1) Determine monitoring sites and analyze the sources and generation sites of hydrodynamic noise in order to simulate the generation of hydrodynamic noise;

[0036] 2) Acquire hydrodynamic noise data at the monitoring sites, filter the hydrodynamic noise data, and extract features. Use the extracted time-frequency domain features as the characterization parameters of the acoustic signal.

[0037] 3) Based on the extracted time-frequency domain features, a random forest is used to construct a leak identification and early warning model to determine whether the fluid flow inside the pipe is abnormal, so as to identify and warn of abnormal states.

[0038] In step 1) above, the monitoring sites are selected based on the upstream and downstream relationship of the water flow in the pipeline network, selecting the upstream hydrodynamic noise excitation points and the locations near the source points of the hydrodynamic noise at elbows and / or tees.

[0039] In this embodiment, the source of acoustic vibration signals collected from the water supply network is analyzed by simulating the fluid flow state within the network. A variety of acoustic signals can be collected from water supply pipelines, among which noise generated by the friction between turbulent fluid and the pipe wall is defined as hydrodynamic noise. Hydrodynamic noise typically originates at pipe fittings in the water supply network, such as elbows, tees, and partially open valves. The fluid within the pipeline is in a flowing state; during this flow, it is impeded at pipe fittings, resulting in turbulence. The interaction between turbulence and the pipe wall generates hydrodynamic noise. By simulating the fluid state at pipe fittings, the fluid pressure and velocity at different locations are revealed, reflecting different intensities of hydrodynamic noise.

[0040] The generation of hydrodynamic noise is simulated, and its theoretical characteristics are analyzed, taking the hydrodynamic noise at a bend as an example. CFD (Computational Fluid Dynamics) is used for simulation, and the simulation of the research object is completed after steps including geometry construction, mesh generation, computation and solution, and post-processing. Figure 2 As shown, the simulation results display the fluid velocity and pressure distribution at the bend. Analysis reveals that the hydrodynamic noise at the bend is velocity-dependent, originating from the friction between fluid particles and the pipe wall. Due to the relatively weak turbulence, the generated vibrational sound waves are weak and low-frequency signals. Experimental results also show the same trend; compared to the vibrational sound waves at the leak point, the hydrodynamic noise source has lower intensity and a lower frequency distribution, such as... Figure 3 The hydrodynamic noise spectrum at the bend shown.

[0041] In step 2) above, hydrodynamic noise data of the monitoring site is acquired, and noise filtering processing is performed on the hydrodynamic noise data, including the following steps:

[0042] 2.1) The acoustic monitoring sensors are deployed at monitoring sites near the pipe fittings to collect hydrodynamic noise data under different conditions concentrated in the low frequency band; the low frequency band is below 500Hz.

[0043] 2.2) Noise filtering is performed on hydrodynamic noise data under different scenarios to remove pipeline interference noise that does not meet the characteristics of hydrodynamic noise, and a hydrodynamic noise dataset is constructed.

[0044] In this embodiment, hydrodynamic noise is generated at the pipe fittings and propagates along the pipe, with the maximum vibration intensity occurring at the pipe fittings. Therefore, the monitoring points for hydrodynamic noise are located near the pipe fittings. Furthermore, since the generation of hydrodynamic noise is related to the relative flow state of the fluid in the pipe, monitoring points located upstream are more likely to cover changes in the downstream flow state; therefore, vibration sensors are deployed upstream of the pipe.

[0045] Specifically, a hydrodynamic noise dataset was collected. Hydrodynamic noise is closely related to the turbulence state within the pipe and also to the type of pipe fittings. This dataset aims to enrich our understanding of hydrodynamic noise. The dataset includes hydrodynamic noise under different conditions, such as pressure, flow velocity, and fitting type. Because hydrodynamic noise is a relatively weak signal, its collection points need to be deployed at specific pipe fitting locations. When a leak occurs downstream, the change in flow state can be transmitted to the upstream location, allowing the corresponding hydrodynamic noise to be collected. Figure 4 As shown, when a leak occurs in the tee-elbow structure, the leakage state at the downstream location is transmitted to the tee-elbow structure through fluid flow. Under this flow state, hydrodynamic noise different from that under normal conditions is generated.

[0046] In this embodiment, the construction of the hydrodynamic noise dataset involves the following steps: Common pipe fittings in water supply networks include elbows, tees, and valves, each with different hydrodynamic noise characteristics. Hydrodynamic noise data is collected under different scenarios, and the noise data characteristics are analyzed to establish a hydrodynamic noise database, providing data support for leak identification and early warning.

[0047] In this embodiment, wavelet transform is used to filter the acquired hydrodynamic noise data. The principle of wavelet filtering will not be elaborated here; the db2 wavelet is used to process and transform the original signal before feature extraction. For example... Figure 5 The diagram shows the coefficients of each signal component generated during wavelet denoising. A five-level wavelet denoising decomposition is used, with the residual term representing the remaining quantity after decomposition. After denoising, some interference noise and signal fluctuations are removed from the signal.

[0048] Specifically, the acoustic data collected from the water supply network includes various components, such as hydrodynamic noise, leakage noise, and interference noise transmitted to the pipeline. The acoustic data monitored by the sensors is decoupled, and the characteristic components of hydrodynamic noise are extracted and characterized, and the hydrodynamic noise components are analyzed. In this embodiment, wavelet transform is used to analyze the collected leakage signal, separate different signal components, and analyze the principal components of hydrodynamic noise.

[0049] In step 2) above, the time-frequency domain features include parameters such as time-domain energy, time-domain average energy, time-domain maximum energy, time-domain kurtosis, time-domain skewness, frequency-domain shape factor, frequency-domain impulse factor, frequency-domain kurtosis factor, and Mel-frequency cepstral coefficients.

[0050] In this embodiment, the following are the calculation formulas for each parameter:

[0051] Time-domain feature calculation, the signal is a time sequence X = [x1 x2 x3 ... x i ... x N ]; where x NLet x1 represent the Nth sequence segment of signal X. Each sequence segment is a uniformly spaced sample of signal X, from x1 to x2. N This constitutes the entire timing signal X.

[0052] The formula for calculating the time-domain energy E is:

[0053]

[0054] Time-domain average energy E mean Calculation formula:

[0055]

[0056] Time-domain maximum energy E max Calculation formula:

[0057] E max =max[x i ]

[0058] The formula for calculating Kurt(x) in the time domain is as follows:

[0059] Kurt(x) = E[(x-μ)] 4 ] / (E[(x-μ) 2 ]) 2

[0060] Where x represents the time-domain value of the signal, and μ represents the time-domain mean of the signal.

[0061] The formula for calculating the time-domain skewness Skew(x) is as follows:

[0062] Skew(x)=E[(x-μ) 3 ] / (E[(x-μ) 2 ]) 2

[0063] Frequency domain characteristic calculation involves first converting the signal from the time domain to the frequency domain: F = [f1 f2 f3 ... f i ... f M This expression represents the calculated frequency F of signal X in different frequency bands after conversion to the frequency domain. M It is the Mth value in the frequency domain.

[0064] Formula for calculating the frequency domain shape factor FD_shape:

[0065]

[0066] Where M represents the number of frequency domain values, f i This represents the i-th frequency domain value.

[0067] Formula for calculating the frequency domain pulse factor FD_pulse:

[0068]

[0069] Formula for calculating frequency domain kurt factor FD_kurt:

[0070]

[0071] in, This represents the frequency domain mean.

[0072] The Mel-frequency cepstral coefficient calculation method involves transforming the signal onto a Mel-frequency scale and then performing a discrete cosine transform to obtain the Mel-frequency cepstral coefficients. The relationship between Mel frequency and normal frequency is expressed as follows:

[0073]

[0074] Among them, f mel f represents the Mel frequency, and f represents the normal frequency.

[0075] Time-frequency domain features are extracted as signal characterization parameters, and the acquired acoustic signals are compressed using data features.

[0076] In step 3) above, the classification and recognition process uses a random forest model. The basic principle of this model is to linearly combine multiple decision trees to form an ensemble model. In a random forest, each decision tree is independent, and the final result is generated by the equal-weighted voting of each tree. By randomly selecting samples and features to generate different decision trees, and giving each decision tree independent judgment capabilities, the robustness and accuracy of the entire model are enhanced. Figure 6 As shown, the established random forest model determines whether a leakage event has occurred at the monitoring location.

[0077] In this embodiment, a leak identification and early warning model is constructed using random forest based on the extracted time-frequency domain features, including the following steps:

[0078] 3.1) The extracted time-frequency domain features are compressed and reconstructed to obtain the feature information matrix;

[0079] 3.2) Construct a leakage identification and early warning model based on the feature information matrix, and establish multiple independent identification decision trees by adopting a mode of random data combination and random feature extraction;

[0080] 3.3) By comprehensively determining the classification of data based on the identification results of multiple decision trees, it is determined whether a leakage event has occurred at the identified monitoring location.

[0081] During the classification process, the characteristic features corresponding to different types of pipe fittings are considered, mainly elbows, tees, and valve structures.

[0082] In one embodiment of the present invention, a water supply network leakage monitoring and early warning system based on hydrodynamic noise is provided, comprising:

[0083] The data analysis module identifies monitoring sites and analyzes the sources and locations of hydrodynamic noise to simulate the generation of hydrodynamic noise.

[0084] The feature extraction module acquires hydrodynamic noise data from the monitoring site, performs noise filtering on the hydrodynamic noise data, and extracts features, using the extracted time-frequency domain features as the characterization parameters of the acoustic signal.

[0085] The early warning module uses a random forest to construct a leak identification and early warning model based on the extracted time-frequency domain features, and determines whether the fluid flow inside the pipe is abnormal, so as to identify and warn of abnormal conditions.

[0086] In the above embodiments, the monitoring sites are selected based on the upstream and downstream relationship of the water flow in the pipeline network, with the upstream hydrodynamic noise excitation point selected near the source of hydrodynamic noise at bends and / or tees.

[0087] In the above embodiments, acquiring hydrodynamic noise data at monitoring sites and performing noise filtering on the hydrodynamic noise data includes:

[0088] Acoustic monitoring sensors are deployed at monitoring sites near pipe fittings to collect hydrodynamic noise data under different conditions, focusing on the low-frequency band.

[0089] Noise filtering is performed on hydrodynamic noise data under different scenarios to remove pipeline interference noise that does not meet the characteristics of hydrodynamic noise, and a hydrodynamic noise dataset is constructed.

[0090] In this embodiment, wavelet transform is used to perform wavelet noise filtering on the collected hydrodynamic noise data.

[0091] In the above embodiments, the time-frequency domain features include: time-domain energy, time-domain average energy, time-domain maximum energy, time-domain kurtosis, time-domain skewness, frequency-domain shape factor, frequency-domain impulse factor, frequency-domain kurtosis factor, and Mel-frequency cepstral coefficients.

[0092] In the above embodiments, a leakage identification and early warning model is constructed using random forest based on the extracted time-frequency domain features, including:

[0093] The extracted time-frequency domain features are compressed and reconstructed to obtain a feature information matrix;

[0094] A leakage identification and early warning model is constructed based on the feature information matrix. Multiple independent identification decision trees are established by adopting a mode of random data combination and random feature extraction.

[0095] The classification of data is determined by combining the identification results of multiple decision trees, and it is then determined whether a leakage event has occurred at the identified monitoring location.

[0096] The system provided in this embodiment is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0097] In one embodiment of the present invention, a computing device is provided, which can be a terminal and may include: a processor, a communication interface, memory, a display screen, and an input device. The processor, communication interface, and memory communicate with each other via a communication bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs, which, when executed by the processor, implement the methods described in the above embodiments. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals. Wireless communication can be achieved through Wi-Fi, a management network, NFC (Near Field Communication), or other technologies. The display screen can be a liquid crystal display or an e-ink display. The input device can be a touch layer covering the display screen, or buttons, a trackball, or a touchpad mounted on the casing of the computing device, or an external keyboard, touchpad, or mouse. The processor can call logical instructions stored in the memory.

[0098] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0099] In one embodiment of the present invention, a computer program product is provided, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, and when the program instructions are executed by a computer, the computer is able to perform the methods provided in the above-described method embodiments.

[0100] In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided, which stores server instructions that cause a computer to perform the methods provided in the above embodiments.

[0101] The computer-readable storage medium provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0102] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0103] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0104] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring and early warning of leaks in water supply networks based on hydrodynamic noise, characterized in that, include: The monitoring sites were determined, and the sources and generation sites of hydrodynamic noise were analyzed in order to simulate the generation of hydrodynamic noise. Hydrodynamic noise data of monitoring sites is acquired, the hydrodynamic noise data is filtered and feature is extracted, and the extracted time-frequency domain features are used as the characterization parameters of the acoustic signal. Based on the extracted time-frequency domain features, a random forest is used to construct a leak identification and early warning model to determine whether the fluid flow inside the pipe is abnormal, so as to identify and warn of abnormal states. Based on the upstream and downstream relationship of water flow in the pipeline network, the monitoring sites are selected from the upstream hydrodynamic noise excitation points and the locations near the hydrodynamic noise source points of elbows and / or tees. Acquire hydrodynamic noise data from monitoring sites and perform noise filtering on the hydrodynamic noise data, including: Acoustic monitoring sensors are deployed at monitoring sites near pipe fittings to collect hydrodynamic noise data under different conditions, focusing on the low-frequency band. Noise filtering is performed on hydrodynamic noise data under different scenarios to remove pipeline interference noise that does not meet the characteristics of hydrodynamic noise, and a hydrodynamic noise dataset is constructed. Time-frequency domain characteristics include: time-domain energy, time-domain average energy, time-domain maximum energy, time-domain kurtosis, time-domain skewness, frequency-domain shape factor, frequency-domain impulse factor, frequency-domain kurtosis factor, and Mel-frequency cepstral coefficients. Based on the extracted time-frequency domain features, a leak identification and early warning model is constructed using random forest, including: The extracted time-frequency domain features are compressed and reconstructed to obtain a feature information matrix; A leakage identification and early warning model is constructed based on the feature information matrix. Multiple independent identification decision trees are established by adopting a mode of random data combination and random feature extraction. The classification of data is determined by combining the identification results of multiple decision trees, and it is then determined whether a leakage event has occurred at the identified monitoring location.

2. The water supply network leakage monitoring and early warning method based on hydrodynamic noise as described in claim 1, characterized in that, Wavelet transform was used to filter the collected hydrodynamic noise data.

3. A water supply network leakage monitoring and early warning system based on hydrodynamic noise, used to implement the water supply network leakage monitoring and early warning method based on hydrodynamic noise as described in claim 1 or 2, characterized in that, include: The data analysis module identifies monitoring sites and analyzes the sources and locations of hydrodynamic noise to simulate the generation of hydrodynamic noise. The feature extraction module acquires hydrodynamic noise data from the monitoring site, performs noise filtering on the hydrodynamic noise data, and extracts features, using the extracted time-frequency domain features as the characterization parameters of the acoustic signal. The early warning module uses a random forest to construct a leak identification and early warning model based on the extracted time-frequency domain features, and determines whether the fluid flow inside the pipe is abnormal, so as to identify and warn of abnormal conditions.

4. The water supply network leakage monitoring and early warning system based on hydrodynamic noise as described in claim 3, characterized in that, The monitoring sites are selected based on the upstream and downstream relationship of water flow in the pipeline network, with the upstream hydrodynamic noise excitation points being selected, and the locations near the hydrodynamic noise source points of elbows and / or tees being selected.

5. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in claims 1 to 2.

6. A computing device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described in claims 1 to 2.

Images