Water supply network leakage locating method, device, equipment, medium and program product

Abstract

The invention relates to the technical field of municipal engineering, and discloses a water supply network leakage locating method, device and equipment, a medium and a program product, and the method comprises the steps: obtaining a transient flow excitation position of a water supply network; simulating a plurality of transient flow events under different leakage conditions based on the transient flow excitation position to obtain a plurality of transient flow frequency domain response amplitudes corresponding to the transient flow events; training a pre-established deep learning network based on the plurality of transient flow frequency domain response amplitudes to obtain a pipe network leakage positioning model; and obtaining the actually measured transient flow frequency domain response amplitude of the water supply network, and positioning the leakage of the water supply network based on the actually measured transient flow frequency domain response amplitude and the network leakage positioning model. According to the method, the purpose of accurately positioning the leakage of the complex pipe network under various leakage working conditions is achieved, and the problem that the leakage positioning is inaccurate due to the fact that an accurate hydraulic model is difficult to construct in the complex pipe network is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of municipal engineering, and in particular to a method, device, equipment, medium and program product for locating leakage in a water supply network. Background Art

[0002] In the study of leakage detection in water supply networks, the leakage detection method based on discrete metering areas (DMAs) can identify leakage within a DMA, but cannot determine the specific location. Usually, it is necessary to use acoustic methods (such as leak detection rods, leak detection correlators, noise recorders, etc.) to accurately locate the area, but this will consume a lot of manpower and time to perform acoustic detection on the entire area. Therefore, there is an urgent need for a method to narrow the search range of acoustic detection to make up for the shortcomings between the two. In this context, the leakage location detection method based on transient flow frequency domain response function shows good application prospects.

[0003] At present, the research on leakage detection methods based on transient flow frequency domain response functions has the following limitations: (1) The research is limited to simple pipeline systems (such as single pipelines, series pipelines, branched or ring pipelines), and the leakage location of complex pipelines is inaccurate; (2) This method is highly dependent on an accurate transient flow hydraulic model, which is not easy to establish.

[0004] Therefore, there is an urgent need for a method that is suitable for accurately locating leakage in complex pipe networks. Summary of the invention

[0005] In view of this, the present invention provides a method, device, equipment, medium and program product for locating leakage in a water supply network to solve the problem of inaccurate leakage positioning caused by difficulty in building an accurate hydraulic model in a complex network.

[0006] In a first aspect, the present invention provides a method for locating leakage in a water supply network, the method comprising:

[0007] Obtain the transient flow excitation position of the water supply network;

[0008] Based on the transient flow excitation position, transient flow events under various leakage conditions are simulated to obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events;

[0009] The pre-established deep learning network is trained based on multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model;

[0010] The measured transient flow frequency domain response amplitude of the water supply network is obtained, and the water supply network leakage is located based on the measured transient flow frequency domain response amplitude and the network leakage location model.

[0011] The water supply network leakage locating method provided by the present invention simulates transient flow events under various leakage conditions based on the transient flow excitation position to obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events; trains a pre-established deep learning network based on the multiple transient flow frequency domain response amplitudes to obtain a network leakage locating model; obtains the measured transient flow frequency domain response amplitudes of the water supply network, and locates the water supply network leakage based on the measured transient flow frequency domain response amplitudes and the network leakage locating model, and uses a deep learning algorithm in a complex network with hydraulic uncertainty to identify the leakage characteristics of the transient flow frequency domain response in the water supply network, thereby determining the pipeline with a higher possibility of leakage, thereby achieving the purpose of accurately locating leakage in a complex network under various leakage conditions, and solving the problem of inaccurate leakage locating caused by the difficulty in constructing an accurate hydraulic model in a complex network.

[0012] In an optional implementation, transient flow events under various leakage conditions are simulated based on the transient flow excitation position, and multiple transient flow frequency domain response amplitudes corresponding to the transient flow events are obtained, including:

[0013] Establish a model network corresponding to the water supply network through mathematical simulation;

[0014] Perform random error correction on the model pipe network, and add leakage parameters to the corrected model pipe network to obtain a leakage model pipe network;

[0015] The transient flow simulation software is used to simulate transient flow events under various leakage conditions of the leakage model pipeline network, and multiple transient flow frequency domain response amplitudes corresponding to the transient flow events under various leakage conditions are calculated.

[0016] The water supply network leakage locating method provided by the present invention establishes a model network corresponding to the water supply network through mathematical simulation, converts the actual water supply network into the model network for research, simplifies the calculation process, performs random error correction on the model network, and adds leakage parameters to the corrected model network to obtain the leakage model network, so that the error of the model network is further reduced, and it is consistent with the characteristics of the actual water supply network, and the simulation effect is good. By using transient flow simulation software to simulate transient flow events under various leakage conditions of the leakage model network, and calculating multiple transient flow frequency domain response amplitudes corresponding to the transient flow events under various leakage conditions, accurate training data is provided for the subsequent training of the deep learning network.

[0017] In an optional embodiment, the leakage parameters include the pipeline where the leakage occurs, the location of the pipeline where the leakage occurs, and the equivalent orifice outflow coefficient;

[0018] The random error of the model pipe network is corrected, and the leakage parameters are added to the corrected model pipe network to obtain the leakage model pipe network including:

[0019] Divide the model pipe network into multiple pipe segments;

[0020] Obtain the equivalent friction coefficient of all pipe sections in the model pipe network, and add a random error that obeys natural distribution to the equivalent friction coefficient to correct the model pipe network;

[0021] Obtain the transient flow wave velocity of all pipe sections in the model pipe network, and add random errors that obey natural distribution to the transient flow wave velocity to correct the model pipe network;

[0022] Using uniform distribution, randomly determine the leaking pipes in the network;

[0023] Use uniform distribution to randomly determine the location of leakage on the pipeline where leakage occurs; the location of leakage is represented by the distance of the leakage from the starting node of the pipeline;

[0024] The equivalent orifice outflow coefficient of the leakage location is randomly determined from a uniform distribution;

[0025] The leaking pipeline, the location of the leaking pipeline and the equivalent orifice outflow coefficient are added to the model pipeline network to obtain the leakage model pipeline network.

[0026] The water supply network leakage locating method provided by the present invention improves the simulation accuracy of the model network by adding naturally distributed random errors to correct the model network, uses uniform distribution to determine the leaking pipes, leakage locations and equivalent orifice outflow coefficients in the network and adds the above parameters to the model network, ensuring that the transient flow frequency domain response amplitude calculated by the model network can cover the characteristics of all pipelines, thereby improving the accuracy of the transient flow frequency domain response amplitude calculation.

[0027] In an optional implementation, before training the pre-established deep learning network based on multiple transient flow frequency domain response amplitudes, the water supply network leakage locating method further includes:

[0028] A plurality of transient flow frequency domain response amplitudes are normalized to obtain a plurality of normalized transient flow frequency domain response amplitudes.

[0029] The water supply network leakage locating method provided by the present invention normalizes multiple transient flow frequency domain response amplitudes to obtain multiple normalized transient flow frequency domain response amplitudes, thereby ensuring the unification of multiple transient flow frequency domain response amplitude data and improving the subsequent training efficiency of the deep learning network.

[0030] In an optional implementation, a pre-established deep learning network is trained based on multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model including:

[0031] Use cross entropy loss as the loss function for a pre-built deep learning network;

[0032] Based on the loss function and multiple normalized transient flow frequency domain response amplitudes, the pre-established deep learning network is trained using the gradient descent method to obtain the pipeline leakage location model.

[0033] The water supply network leakage locating method provided by the present invention uses cross entropy loss as the loss function of a pre-established deep learning network; based on the loss function and multiple normalized transient flow frequency domain response amplitudes, the pre-established deep learning network is trained using the gradient descent method to obtain a network leakage locating model; with the help of the deep learning network, the network leakage characteristics in the transient flow frequency domain response are effectively and fully mined, and the influence of the hydraulic model error on the leakage locating result is reduced; the obtained network leakage locating model provides conditions for subsequent leakage locating of the actual water supply network.

[0034] In an optional implementation, obtaining the measured transient flow frequency domain response amplitude of the water supply network, and locating the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage location model includes:

[0035] Selecting a transient flow excitation position of an actual water supply network, and calculating a measured transient flow frequency domain response amplitude based on the transient flow excitation position of the actual water supply network;

[0036] Normalizing the measured transient flow frequency domain response amplitude to obtain the normalized measured transient flow frequency domain response amplitude;

[0037] The normalized measured transient flow frequency domain response amplitude is input into the pipe network leakage location model, and the water supply pipe network leakage location result is output.

[0038] The water supply network leakage locating method provided by the present invention adopts a network leakage locating model obtained based on a deep learning network to identify the leakage characteristics of the transient flow frequency domain response in the network, thereby determining the pipeline with a higher possibility of leakage, thereby achieving the purpose of locating the leakage of the actual water supply network.

[0039] In a second aspect, the present invention provides a water supply network leakage locating device, the device comprising:

[0040] A position acquisition module, used to acquire the transient flow excitation position of the water supply network;

[0041] A transient flow frequency domain response amplitude calculation module is used to simulate transient flow events under various leakage conditions based on transient flow excitation positions, and obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events;

[0042] A training module is used to train a pre-established deep learning network based on multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model;

[0043] The water supply network leakage locating module is used to obtain the measured transient flow frequency domain response amplitude of the water supply network, and locate the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage locating model.

[0044] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the water supply network leakage locating method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0045] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for locating leakage in a water supply network according to the first aspect or any corresponding embodiment thereof.

[0046] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the method for locating leakage in a water supply network according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 is a schematic flow chart of a method for locating leakage in a water supply network according to an embodiment of the present invention;

[0049] Figure 2 is a schematic flow chart of another method for locating leakage in a water supply network according to an embodiment of the present invention;

[0050] Figure 3 is a schematic flow chart of another method for locating leakage in a water supply network according to an embodiment of the present invention;

[0051] Figure 4 is a schematic diagram of the actual water supply network distribution according to an embodiment of the present invention;

[0052] Figure 5 is a structural block diagram of a water supply network leakage locating device according to an embodiment of the present invention;

[0053] Figure 6 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0055] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0056] According to an embodiment of the present invention, an embodiment of a method for locating leakage in a water supply network is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0057] In this embodiment, a method for locating water supply network leakage is provided, which can be used in mobile terminals, such as central processing units, servers, etc. Figure 1 is a flow chart of a method for locating water supply network leakage according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0058] Step S101, obtaining a transient flow excitation position of a water supply network.

[0059] Specifically, transient flow is when the water flow state changes from one stable state to another stable state, and the unstable flow state in the middle is called the transition process. The transient flow excitation position refers to the position close to the critical pressure bearing capacity of the pipeline. The selection of the transient flow excitation position mainly depends on whether it is convenient to implement the test in the actual project, that is, the transient flow excitation position is selected according to the actual project, and no specific limitation is made here.

[0060] Step S102 , simulating transient flow events under a plurality of different leakage conditions based on the transient flow excitation position, and obtaining a plurality of transient flow frequency domain response amplitudes corresponding to the transient flow events.

[0061] Specifically, when the transient flow excitation position is obtained, the transient flow event will be generated by the pulse operation of a controllable valve that flows out to the atmosphere. The transient flow events under various leakage conditions are simulated by transient flow simulation software, and the transient flow frequency domain response amplitude under various pipe network burst positions, burst sizes and pipe network model parameters are obtained.

[0062] Step S103: training a pre-established deep learning network based on multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model.

[0063] Specifically, the deep learning network of this embodiment is composed of several densely connected blocks, several conversion layers and a classifier.

[0064] Each densely connected block contains several aggregation functions. The aggregation function is a composite function, each of which consists of batch normalization (BN), rectified linear unit (ReLU) and linear connection (Linear). The i-th aggregation function in each densely connected block is denoted as L i , L i Instead of just outputting the tensor x i is not passed to the next aggregate function, but to all subsequent aggregate functions, that is:

[0065] x i =L i ([x0, x1, ..., x i-1 [) (1);

[0066] Among them, x0 is the input tensor of the densely connected block, [x0, x1, ..., x i-1 ] represents the tensor x0, x1, ..., x i-1 If the length of the output tensor of each aggregation function is k, then the i-th aggregation function L i The length of the input tensor is k0+k(i-1), where k0 is the length of the tensor x0 of the input densely connected block, and the hyperparameter k is called the growth rate.

[0067] Each conversion layer is composed of BN, ReLU and Linear. The conversion layer is responsible for receiving the output of the previous densely connected block, compressing the length of the output, and then passing it to the next densely connected block. If the input tensor of the conversion layer (that is, the output tensor of the previous densely connected block) is m, the length of its output tensor should be in, It means rounding down, and the hyperparameter 0<θ≤1 is called the compression factor.

[0068] The classifier contains a BN-ReLU-Linear module and a LogSoftmax function (data normalization function). If the total number of pipes in the water supply network is p, the BN-ReLU-Linear module in the classifier receives the output of the last densely connected block and passes a tensor of length p to the LogSoftmax function. For a tensor of length p, Z = [Z1, ..., Z p], its i-th LogSoftmax function value σ i for:

[0069]

[0070] Output tensor Z = [Z1, ..., Z p ] is a sequence of leakage judgment values. The larger a value in the sequence is, the greater the possibility of leakage in the pipeline corresponding to the value is.

[0071] Multiple transient flow frequency domain response amplitudes are input into the established deep learning network, and the deep learning network is trained with multiple transient flow frequency domain response amplitudes to obtain the pipeline leakage location model.

[0072] Step S104, obtaining the measured transient flow frequency domain response amplitude of the water supply network, and locating the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage location model.

[0073] Specifically, the measured transient flow frequency domain response amplitude of the water supply network is obtained, and the measured transient flow frequency domain response amplitude is input into the network leakage location model to obtain the leakage location result.

[0074] The method for locating leakage in a water supply network provided in the present embodiment simulates transient flow events under various leakage conditions based on the transient flow excitation position to obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events; trains a pre-established deep learning network based on the multiple transient flow frequency domain response amplitudes to obtain a network leakage locating model; obtains the measured transient flow frequency domain response amplitudes of the water supply network, and locates the leakage in the water supply network based on the measured transient flow frequency domain response amplitudes and the network leakage locating model; uses a deep learning algorithm in a complex network with hydraulic uncertainty to identify the leakage characteristics of the transient flow frequency domain response in the water supply network, thereby determining the pipeline with a higher possibility of leakage, thereby achieving the purpose of accurately locating leakage in a complex network under various leakage conditions, and solving the problem of inaccurate leakage locating caused by the difficulty in constructing an accurate hydraulic model in a complex network.

[0075] In this embodiment, a method for locating water supply network leakage is provided, which can be used in mobile terminals, such as central processing units, servers, etc. Figure 2 is a flow chart of a method for locating water supply network leakage according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0076] Step S201, obtain the transient flow excitation position of the water supply network. Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0077] Step S202 , simulating transient flow events under a plurality of different leakage conditions based on the transient flow excitation position, and obtaining a plurality of transient flow frequency domain response amplitudes corresponding to the transient flow events.

[0078] Specifically, the above step S202 includes:

[0079] Step S2021, establishing a model network corresponding to the water supply network through mathematical simulation.

[0080] Specifically, the mathematical simulation method may use mathematical simulation software, such as Hammer software, and Hammer software may be used to establish a model network corresponding to the actual water supply network.

[0081] Step S2022, performing random error correction on the model pipe network, and adding leakage parameters to the corrected model pipe network to obtain a leakage model pipe network.

[0082] Since there is an error between the model network obtained by mathematical simulation and the actual water supply network, the model network needs to be corrected to ensure the consistency between the model network and the actual water supply network. And by adding leakage parameters to the model network, the model network can cover the leakage characteristics of all networks.

[0083] In some optional implementations, the leakage parameters include the leakage pipeline, the leakage pipeline location, and the equivalent orifice outflow coefficient. The above step S2022 includes:

[0084] Step a1: divide the model pipe network into multiple pipe sections.

[0085] Step a2, obtaining the equivalent friction coefficients of all pipe sections in the model pipe network, and adding random errors that obey natural distribution to the equivalent friction coefficients to correct the model pipe network.

[0086] Specifically, the equivalent friction coefficient is also called the Darcy-Weisbach friction coefficient, and its natural distribution is Add the equivalent Darcy-Weisbach friction coefficient of all pipes in the model network to the distribution The random error of f Represents standard deviation.

[0087] Step a3, obtaining the transient flow wave velocity of all pipe sections in the model pipe network, and adding a random error that obeys natural distribution to the transient flow wave velocity to correct the model pipe network.

[0088] Specifically, the natural distribution is Add a distribution-dependent transient velocity to all pipes in the model network. The random error of a Represents standard deviation.

[0089] Step a4, using uniform distribution, randomly determine the leaking pipeline in the pipeline network.

[0090] Specifically, the uniform distribution is a discrete uniform distribution. When determining the leaking pipeline, the discrete uniform distribution is used so that the probability of any pipeline in the pipeline network being selected as the leakage occurrence pipeline is equal.

[0091] Step a5, using uniform distribution, randomly determine the leakage location on the pipeline where the leakage occurs; the leakage location is represented by the distance of the leakage from the starting node of the pipeline.

[0092] Specifically, the uniform distribution is U(0, L), where L is the length of the leaking pipeline. Using the uniform distribution, the location where the leak occurs on the pipeline is randomly determined, that is, the distance of the leak from the starting node of the pipeline.

[0093] Step a6, randomly determining the equivalent orifice outflow coefficient of the leakage location from the uniform distribution.

[0094] Specifically, from the uniform distribution U(C min , C max ), the equivalent orifice outflow coefficient C at the leakage location is randomly determined, and the residual water head H at the pipe burst location is L and flow rate Q L obey:

[0095]

[0096] Among them, C min , C max They represent the minimum and maximum values ​​of the equivalent orifice discharge coefficient respectively.

[0097] Step a7, adding the leaking pipeline, the location of the leaking pipeline and the equivalent orifice outflow coefficient to the model pipeline network to obtain the leakage model pipeline network.

[0098] The water supply network leakage locating method provided in this embodiment improves the simulation accuracy of the model network by adding naturally distributed random errors to correct the model network, uses uniform distribution to determine the leaking pipes, leakage locations and equivalent orifice outflow coefficients in the network and adds the above parameters to the model network, ensuring that the transient flow frequency domain response amplitude calculated by the model network can cover the characteristics of all pipes, thereby improving the accuracy of the transient flow frequency domain response amplitude calculation.

[0099] Step S2023, using transient flow simulation software to simulate transient flow events under various leakage conditions of the leakage model pipeline network, and calculating multiple transient flow frequency domain response amplitudes corresponding to the transient flow events under various leakage conditions.

[0100] Specifically, the transient flow frequency domain response amplitude F of the pipe network is defined as follows:

[0101]

[0102] in, and They are the frequency domain water head disturbance and frequency domain outflow disturbance of the controllable valve, which are obtained by Fourier transforming the time domain water head disturbance h and the time domain outflow disturbance q respectively.

[0103] Among them, the time domain head disturbance h is:

[0104] h=H-H0 (5);

[0105] Among them, H is the time series of the water head at the valve after the transient flow event is excited, and H0 is the water head value at the valve before the transient flow event is excited.

[0106] Also available:

[0107] q=Q-Q0 (6);

[0108] Among them, Q is the outflow time series at the valve after the transient flow event is triggered, and Q0 is the outflow at the valve before the transient flow event is triggered.

[0109] Use transient flow simulation software to simulate the excited transient flow events, solve the transient flow frequency domain response amplitude of the leakage model pipe network, record the transient flow frequency domain response amplitude and the corresponding leakage pipe. Repeat steps S2022 and S2023 several times to obtain multiple transient flow frequency domain response amplitudes corresponding to transient flow events under multiple different leakage conditions.

[0110] Step S203: training the pre-established deep learning network based on multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0111] Step S204, obtaining the measured transient flow frequency domain response amplitude of the water supply network, and locating the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage location model. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0112] The water supply network leakage locating method provided in this embodiment establishes a model network corresponding to the water supply network through mathematical simulation, converts the actual water supply network into the model network for research, simplifies the calculation process, performs random error correction on the model network, and adds leakage parameters to the corrected model network to obtain the leakage model network, so that the error of the model network is further reduced, and is consistent with the characteristics of the actual water supply network, with good simulation effect. By using transient flow simulation software to simulate transient flow events under various leakage conditions of the leakage model network, and calculating multiple transient flow frequency domain response amplitudes corresponding to the transient flow events under various leakage conditions, accurate training data is provided for the subsequent training of the deep learning network.

[0113] In this embodiment, a method for locating water supply network leakage is provided, which can be used in mobile terminals, such as central processing units, servers, etc. Figure 3 is a flow chart of a method for locating water supply network leakage according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0114] Step S301, obtain the transient flow excitation position of the water supply network. Figure 2 Step S201 of the illustrated embodiment will not be described in detail here.

[0115] Step S302: Simulate transient flow events under various leakage conditions based on the transient flow excitation position to obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events. Figure 2 Step S201 of the illustrated embodiment will not be described in detail here.

[0116] Step S303 , normalizing the multiple transient flow frequency domain response amplitudes to obtain multiple normalized transient flow frequency domain response amplitudes.

[0117] Specifically, the transient flow frequency domain response amplitude is normalized from the minimum to the maximum value, and the normalization formula is:

[0118]

[0119] in, is the normalized transient flow frequency domain response amplitude vector, F min and F max are respectively the minimum and maximum values ​​in the original transient flow frequency domain response amplitude vector F.

[0120] The data recorded in step S303 are organized into a data set, the label of the data set is the pipeline where the leakage occurs, and the corresponding data is the normalized transient flow frequency domain response amplitude.

[0121] Step S304: training a pre-established deep learning network based on multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model.

[0122] Specifically, the above step S304 includes:

[0123] Step S3041, using the cross entropy loss as the loss function of the pre-established deep learning network.

[0124] Specifically, cross entropy loss is a loss function commonly used in machine learning and deep learning, and is particularly suitable for classification tasks, and is used to measure the difference between the prediction results of the model and the true label. Cross entropy loss is used as the loss function of the deep learning network in this embodiment.

[0125] Step S3042: Based on the loss function and multiple normalized transient flow frequency domain response amplitudes, a pre-established deep learning network is trained using a gradient descent method to obtain a pipeline network leakage location model.

[0126] Specifically, gradient descent is a first-order optimization algorithm that is used to recursively approximate the minimum deviation. To use gradient descent to find the local minimum of a function, the model must iteratively search for points at a specified step distance in the opposite direction of the gradient (or approximate gradient) corresponding to the current point on the function.

[0127] Cross Entropy Loss is used as the loss function, and the gradient descent method and multiple normalized transient flow frequency domain response amplitudes are used to train the deep learning network. After a certain number of repetitions, the training is completed and the pipeline leakage location model is obtained.

[0128] Step S305, obtaining the measured transient flow frequency domain response amplitude of the water supply network, and locating the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage location model.

[0129] Specifically, the above step S305 includes:

[0130] Step S3051, selecting a transient flow excitation position of an actual water supply network, and calculating a measured transient flow frequency domain response amplitude based on the transient flow excitation position of the actual water supply network.

[0131] Specifically, a field experiment is conducted at a selected transient flow excitation position of an actual water supply network to obtain multiple transient flow frequency domain response amplitudes of the actual water supply network in the experiment.

[0132] Step S3052, normalizing the measured transient flow frequency domain response amplitude to obtain the normalized measured transient flow frequency domain response amplitude.

[0133] Specifically, the minimum-maximum normalization formula (7) is used to normalize the measured transient flow frequency domain response amplitude to obtain the normalized measured transient flow frequency domain response amplitude.

[0134] Step S3053, input the normalized measured transient flow frequency domain response amplitude into the pipe network leakage location model, and output the water supply pipe network leakage location result.

[0135] Specifically, the normalized measured transient flow frequency domain response amplitude is input into the pipeline network leakage location model to obtain the leakage location result.

[0136] The water supply network leakage location method provided in this embodiment normalizes multiple transient flow frequency domain response amplitudes to obtain multiple normalized transient flow frequency domain response amplitudes, ensuring the unification of multiple transient flow frequency domain response amplitude data and improving the subsequent training efficiency of the deep learning network. The cross entropy loss is used as the loss function of the pre-established deep learning network; the pre-established deep learning network is trained using the gradient descent method based on the loss function and multiple normalized transient flow frequency domain response amplitudes to obtain a network leakage location model. With the help of the deep learning network, the network leakage characteristics in the transient flow frequency domain response are effectively and fully mined, and the influence of the hydraulic model error on the leakage location results is reduced. The network leakage location model obtained based on the deep learning network is used to identify the leakage characteristics of the transient flow frequency domain response in the network, thereby determining the pipelines with a higher possibility of leakage, and achieving the purpose of locating the leakage of the actual water supply network.

[0137] As one or more specific application embodiments of the present invention, Figure 4 The water supply network leakage locating method provided by the present invention is further described in detail as follows:

[0138] like Figure 4 As shown, the water supply of the pipe network is 14400m 3 Every day, there are 127 outflow nodes, 149 pipes, and 1 pool. The total length of the pipes is 45.8 km. The transient flow excitation locations are as follows: Figure 4 Indicated by the middle triangle.

[0139] The deep learning network in this embodiment includes 4 densely connected blocks, each of which includes 8, 16, 16, and 12 aggregation functions, respectively. The growth rate k = 128, and the compression factor θ = 1 / 3. The initial learning rate is 0.6, and the learning rate is updated to 1 / 3 of the original after every 20 generations of training.

[0140] The following parameters are used for simulation: f =0.005,σ a =50m / s, Cmin =7.534×10 -4 m 2.5 s -1 , C max =15.068×10 -4 m 2.5 s -1 . The sampling frequency of transient flow monitoring data is 125Hz, and the sampling time is 400s. The deep learning network is trained using the data set generated by the model, and then the measured data is input into the trained deep learning network, i.e., the pipe network leakage location model, and the five pipes with the highest possibility of leakage are returned. The results show that the pipe that actually burst is included in the five returned pipes. The water supply pipe network leakage location method provided by the present invention can effectively locate the leakage position in the pipe network and determine the pipes with a higher possibility of leakage.

[0141] In this embodiment, a water supply network leakage locating device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0142] This embodiment provides a water supply network leakage locating device, such as Figure 5 As shown, including:

[0143] The position acquisition module 501 is used to acquire the transient flow excitation position of the water supply network.

[0144] The transient flow frequency domain response amplitude calculation module 502 is used to simulate transient flow events under various leakage conditions based on transient flow excitation positions, and obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events.

[0145] The training module 503 is used to train a pre-established deep learning network based on multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model.

[0146] The water supply network leakage locating module 504 is used to obtain the measured transient flow frequency domain response amplitude of the water supply network, and locate the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage locating model.

[0147] In some optional implementations, the water supply network leakage locating device further includes:

[0148] The normalization module is used to perform normalization processing on multiple transient flow frequency domain response amplitudes to obtain multiple normalized transient flow frequency domain response amplitudes.

[0149] In some optional implementations, the transient flow frequency domain response amplitude calculation module 502 includes:

[0150] The model pipe network establishment unit is used to establish a model pipe network corresponding to the water supply pipe network through mathematical simulation.

[0151] The model correction and parameter addition unit is used to correct the random errors of the model pipe network and add the leakage parameters to the corrected model pipe network to obtain the leakage model pipe network.

[0152] The simulation unit is used to simulate transient flow events under various leakage conditions of the leakage model pipeline network using transient flow simulation software, and calculate multiple transient flow frequency domain response amplitudes corresponding to the transient flow events under various leakage conditions.

[0153] In some optional embodiments, the leakage parameters include the leakage pipeline, the leakage pipeline location and the equivalent orifice outflow coefficient, and the model correction and parameter addition unit includes:

[0154] The pipe segment division subunit is used to divide the model pipe network into multiple pipe segments.

[0155] The first correction subunit is used to obtain the equivalent friction coefficients of all pipe sections in the model pipe network, and to add random errors that obey natural distribution to the equivalent friction coefficients to correct the model pipe network.

[0156] The second correction subunit is used to obtain the transient flow wave velocity of all pipe sections in the model pipe network, and add a random error that obeys the natural distribution to the transient flow wave velocity to correct the model pipe network.

[0157] The leaking pipeline determination unit is used to randomly determine the leaking pipelines in the pipeline network using uniform distribution.

[0158] The leakage location determination subunit is used to randomly determine the leakage location on the pipeline where the leakage occurs using uniform distribution; the leakage location is represented by the distance of the leakage from the starting node of the pipeline.

[0159] The equivalent orifice outflow coefficient determination subunit is used to randomly determine the equivalent orifice outflow coefficient of the leakage location from a uniform distribution.

[0160] The parameter adding subunit is used to add the leaking pipeline, the location of the leaking pipeline and the equivalent orifice outflow coefficient to the model pipe network to obtain the leakage model pipe network.

[0161] In some optional implementations, the training module 503 includes:

[0162] A loss function determination unit is used to use the cross entropy loss as the loss function of the pre-established deep learning network.

[0163] The training unit is used to train a pre-established deep learning network using a gradient descent method based on a loss function and multiple normalized transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model.

[0164] In some optional implementations, the water supply network leakage locating module 504 includes:

[0165] The measured transient flow frequency domain response amplitude determination unit is used to select the transient flow excitation position of the actual water supply network, and calculate the measured transient flow frequency domain response amplitude based on the transient flow excitation position of the actual water supply network.

[0166] The data normalization unit is used to perform normalization processing on the measured transient flow frequency domain response amplitude to obtain the normalized measured transient flow frequency domain response amplitude.

[0167] The positioning unit is used to input the normalized measured transient flow frequency domain response amplitude into the pipe network leakage positioning model and output the water supply pipe network leakage positioning result.

[0168] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0169] The water supply network leakage locating device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0170] The embodiment of the present invention also provides a computer device having the above Figure 5 The water supply network leakage locating device shown.

[0171] See also Figure 6 , Figure 6 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 6As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 6 A processor 10 is taken as an example.

[0172] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0173] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0174] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0175] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0176] The computer device also includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Figure 6 The example of connecting through bus is taken in the following.

[0177] The input device 30 can receive input digital or character information, and generate key signal input related to the user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator bar, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light emitting diode, a display and a plasma display. In some optional embodiments, the display device can be a touch screen.

[0178] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0179] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0180] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for locating leakage in a water supply network, characterized in that: The method comprises: Obtain the transient flow excitation position of the water supply network; Simulating transient flow events under various leakage conditions based on the transient flow excitation position to obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events; Based on the multiple transient flow frequency domain response amplitudes, a pre-established deep learning network is trained to obtain a pipeline network leakage location model; The measured transient flow frequency domain response amplitude of the water supply network is obtained, and the leakage of the water supply network is located based on the measured transient flow frequency domain response amplitude and the network leakage location model.

2. The method according to claim 1, characterized in that Based on the transient flow excitation position, transient flow events under various leakage conditions are simulated to obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events, including: Establish a model network corresponding to the water supply network through mathematical simulation; Performing random error correction on the model pipe network, and adding leakage parameters to the corrected model pipe network to obtain a leakage model pipe network; The transient flow simulation software is used to simulate transient flow events under various leakage conditions of the leakage model pipeline network, and multiple transient flow frequency domain response amplitudes corresponding to the transient flow events under various leakage conditions are calculated.

3. The method according to claim 2, characterized in that The leakage parameters include the leakage pipeline, the leakage pipeline location and the equivalent orifice outflow coefficient; The random error correction is performed on the model pipe network, and the leakage parameter is added to the corrected model pipe network to obtain the leakage model pipe network, including: Divide the model pipe network into multiple pipe segments; Obtaining the equivalent friction coefficients of all pipe sections in the model pipe network, and adding random errors that obey natural distribution to the equivalent friction coefficients to correct the model pipe network; Obtaining the transient flow wave velocity of all pipe sections in the model pipe network, and adding a random error that obeys natural distribution to the transient flow wave velocity to correct the model pipe network; Using uniform distribution, randomly determine the leaking pipes in the network; Using uniform distribution, randomly determine the leakage location on the leaking pipeline; the leakage location is represented by the distance of the leakage from the starting node of the pipeline; Randomly determine the equivalent orifice outflow coefficient of the leakage location from a uniform distribution; The leaking pipeline, the location of the leaking pipeline and the equivalent orifice outflow coefficient are added to the model pipeline network to obtain the leakage model pipeline network.

4. The method according to claim 1, characterized in that: Before training the pre-established deep learning network based on the multiple transient flow frequency domain response amplitudes, the method further includes: The multiple transient flow frequency domain response amplitudes are normalized to obtain multiple normalized transient flow frequency domain response amplitudes.

5. The method according to claim 4, characterized in that The pre-established deep learning network is trained based on the multiple transient flow frequency domain response amplitudes to obtain a pipeline network leakage location model including: Use cross entropy loss as the loss function for a pre-built deep learning network; Based on the loss function and multiple normalized transient flow frequency domain response amplitudes, a pre-established deep learning network is trained using a gradient descent method to obtain a pipeline network leakage location model.

6. The method according to claim 1, characterized in that Obtaining the measured transient flow frequency domain response amplitude of the water supply network, and locating the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage location model includes: Selecting a transient flow excitation position of an actual water supply network, and calculating a measured transient flow frequency domain response amplitude based on the transient flow excitation position of the actual water supply network; Normalizing the measured transient flow frequency domain response amplitude to obtain a normalized measured transient flow frequency domain response amplitude; The normalized measured transient flow frequency domain response amplitude is input into the pipe network leakage location model, and the water supply pipe network leakage location result is output.

7. A water supply network leakage locating device, characterized in that: The device comprises: A position acquisition module, used to acquire the transient flow excitation position of the water supply network; A transient flow frequency domain response amplitude calculation module is used to simulate transient flow events under various leakage conditions based on the transient flow excitation position to obtain multiple transient flow frequency domain response amplitudes corresponding to the transient flow events; A training module, used to train a pre-established deep learning network based on the multiple transient flow frequency domain response amplitudes to obtain a pipe network leakage location model; The water supply network leakage locating module is used to obtain the measured transient flow frequency domain response amplitude of the water supply network, and locate the water supply network leakage based on the measured transient flow frequency domain response amplitude and the network leakage locating model.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the water supply network leakage locating method according to any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the water supply network leakage locating method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions, wherein the computer instructions are used to enable a computer to execute the water supply network leakage locating method according to any one of claims 1 to 6.

Images