A method and system for evaluating leakage rate of constant pressure water supply network

Abstract

The application belongs to the technical field of industrial automatic control, and discloses a leakage rate evaluation method and system suitable for constant pressure water supply pipe network. The leakage rate evaluation method suitable for constant pressure water supply pipe network comprises the following steps: according to the characteristics of the constant pressure water supply system, using the Lagrange proportional load method and the Newton iteration method, the water consumption signal reference vector and the leakage signal reference vector are separated from the total table flow signal and the pressure signal of the pressurizing equipment; and taking the separated water consumption signal and the leakage signal as the constraint conditions, the constraint independent component analysis algorithm is combined to carry out iteration, so that the separated water consumption signal and the leakage signal are obtained. Even if the current industrial park, station and other field sections do not have automatic flow and pressure collection remote devices, the application only needs to install a common pressure gauge at the water pump during the shutdown and maintenance time, install a common water gauge at the water pump outlet pipe, and send personnel to record the reading at regular time, so that the collection cost of observation data is extremely low.

Description

Technical Field

[0001] This invention belongs to the field of industrial automatic control technology, and in particular relates to a method and system for assessing leakage rate in constant pressure water supply networks. Background Technology

[0002] Many industrial parks, railway stations, and other facilities with independent water supply systems suffer from poor pipe quality and inconsistent construction quality, leading to gradual aging and severe leakage in some pipe networks and resulting in significant water waste. For example, in one high-speed rail depot, the leakage rate has reached nearly 40%, making pipe replacement urgent. However, replacing the pipe network is extremely costly and will significantly impact the daily water use of the facility during the construction period. Therefore, it should not be undertaken blindly; a thorough assessment of the leakage situation and a careful calculation of the economic costs are essential. Currently, the main leakage detection technologies include the following:

[0003] 1. Instrument testing

[0004] The main components include a leak detection rod, which locates leaks by detecting certain levels of water leakage sound; a landmark radar, which detects underground water leakage via radio waves and displays images of the leak points; and a multi-probe correlator, which combines leak point prediction and precise location, and the testing time is not limited to nighttime and is not affected by other noise interference.

[0005] However, these methods have a certain lag, not only consuming manpower and resources, but also having a long leak detection cycle, and reducing the leakage rate only remains at a superficial level.

[0006] 2. Area-based leak detection method

[0007] The water supply network is divided into zones. For each water supply zone, a small number of water meters are kept at the inlet pipe, and all valves connecting the zone to the outside are closed. Water consumption in each zone is then measured. The difference between the water meter reading at the zone inlet and the readings of all users within the zone represents the leakage rate. This method, through systematic data mining, can be used for network data analysis. Comparing the measured flow rate with the normal flow rate can help detect early signs of leakage. However, this method requires a large initial investment and demands high measurement accuracy. Water usage points in different areas often lack complete sub-metering capabilities, making it difficult to determine the leakage rate based on the difference between the main meter reading and the sub-meter readings.

[0008] 3. Water balance method

[0009] First, online metering is performed based on water meters or flow rates to determine the total water supply Q of the pipeline network. Z The effective water supply volume Q was determined again. a And estimate the apparent leakage F, then the physical leakage Q of the pipe network. L =Q Z -Q a -F. Some data, such as free water supply and apparent water supply, cannot be accurately obtained by this method and are only estimated values, so the calculation results are not rigorous.

[0010] 4. Kalman filter method

[0011] The filter theory has a wide application field after converting the frequency domain variable into the time domain variable. Since 14 years, the Kalman filter has been gradually applied to the evaluation of leakage in water supply networks. The precondition for using the Kalman filter is that the signal has a determined power spectrum. However, only when the minimum unit is a day, the variance of the physical leakage of the pipe network is approximately equal. If a smaller time interval is required to analyze the leakage signal, the prior condition may not be satisfied.

[0012] 5. Blind source separation method

[0013] The blind source separation was first proposed in 1986. Since 13 years, the blind source separation theory has been applied to pipe networks for the first time.

[0014] The blind source separation generally uses the independent component analysis (ICA) algorithm. The CICA algorithm has the widest application and the best effect. Compared with other ICA algorithms, the CICA algorithm introduces a reference vector, changes the model from full blindness to semi-blindness, and has the greatest advantage of not being strict with the independence of the source signal. However, the CICA algorithm has the following disadvantages:

[0015] (1) The selection of the reference vector is very difficult. The CICA algorithm needs to input a reference vector artificially. The accuracy of the reference vector directly affects the separation result of the CICA algorithm. However, in the face of a large underground pipe network, it is often difficult to obtain useful reference information in the absence of complete metering tools. The algorithm has a large number of parameters that need to be adjusted artificially. The values of the parameters greatly affect the separation result.

[0016] (2) The CICA algorithm is prone to false convergence. Many scholars have proposed that the CICA algorithm is prone to false convergence due to the influence of the artificially input parameters. Although some measures have been proposed, these measures cannot completely solve the false convergence problem as the application conditions increase.

[0017] The above two disadvantages are caused by the inaccuracy of the artificially input parameters, and are the problems and defects that the present application focuses on solving. The existing technology has the following problems and defects: (1) In the existing technology, due to the subjectivity of the artificially input parameters, the accuracy of the separated signal is low. (2) In the existing technology, the reference signal of the water consumption cannot be separated from the total metering flow signal and the pressure signal of the pressurizing equipment, so that the accuracy of the reference signal is poor. (3) In the existing technology, the cost of collecting the observation data is high. SUMMARY

[0018] In order to overcome the problems in the related art, the present application discloses a method and system for evaluating the leakage rate of a constant pressure water supply pipe network. Specifically, the present application relates to a CICA algorithm for evaluating the leakage rate of a constant pressure water supply pipe network.

[0019] The technical solution is as follows: A leakage rate evaluation method suitable for constant pressure water supply network includes: according to the characteristics of the constant pressure water supply system, using the Lagrange proportional load method and the Newton iteration method, the water signal reference vector and the leakage signal reference vector are separated from the total meter flow signal and the pressure signal of the pressurizing equipment, and the separated water consumption signal and the leakage signal are obtained by iteration of the constraint independent component analysis algorithm.

[0020] Further, the method for separating the water signal reference vector from the total meter flow signal by using the Lagrange proportional load method and the Newton iteration method includes:

[0021] constructing a Lagrange function E{zG'(w T z)}+λw=0;

[0022] using the Newton method for iteration, and the iteration rule is:

[0023] w i+1 =E{zG'(w i T z)}-E{G”(w i T z)}w i ;

[0024] w i+1 =w i+1 / ||w i+1 ||;

[0025] After iteration, it is judged whether |w i+1 -w i |≤ξ is established or not.

[0026] If |w i+1 -w i |≤ξ is established, the iteration is ended, and the water signal reference vector r1=w i+1 z is output.

[0027] Further, if |w i+1 -w i |≤ξ is not established, the Newton method is used for iteration again.

[0028] Further, the leakage signal reference vector separated from the pressure signal of the pressurizing equipment includes: the pipe network is a constant pressure water supply system, the leakage reference vector is regarded as a constant, and the leakage reference vector is a reference vector directly determined according to the habit of constant pressure water supply, and the leakage reference vector r2=ones(1,length(z)).

[0029] Further, before the water signal reference vector and the leakage signal reference vector are separated from the total meter flow signal and the pressure signal of the pressurizing equipment by using the Lagrange proportional load method and the Newton iteration method, it is necessary to perform:

[0030] (1) Construct a blind source separation model for daily water consumption and leakage in water supply networks:

[0031] x = As

[0032] x represents the observed signal, referring to the flow and pressure signals of the constant pressure water supply equipment. The signals are collected multiple times per hour, and the average value is taken after removing outliers as the signal for that time period. There is one set of signals per hour, for a total of 24 sets per day.

[0033]

[0034] s represents the source signal, the average hourly water consumption and leakage signal, with 24 sets throughout the day;

[0035]

[0036] Create a separation matrix W such that y = Wx = WAs = Ks. When K approaches the identity matrix, the separation signal y = s is achieved.

[0037]

[0038] According to the central limit theorem, Σk i s i Gaussian type ratio of any s i None of them are weak; the Gaussian property of y is weakest if and only if y is one of the elements in s; only k in K has this property. i If the value is non-zero, then y is considered a successfully separated independent component; the Gaussian distribution has the largest entropy value, so we construct an objective function that maximizes the negative entropy of y;

[0039] Objective function: max J(y) = max(E{G(y)} - E{G(y)} guass )}) 2 =(E{G(w T z)}-E{G(y guass )}) 2 ;

[0040] J(y) is the negative entropy of y; H(y) is the entropy of y, y guass G(y) is a Gaussian random vector with the same covariance matrix as y and has maximum entropy; G(y) has different Gaussian signals depending on the degree of Gaussianity of the source signal, including:

[0041] 1) The source signals are all super-Gaussian signals:

[0042]

[0043] 2) The source signals are all sub-Gaussian signals:

[0044] G(y)=0.25y 4G'(y) = y 3 ;

[0045] 3) source signal is super-Gaussian signal and sub-Gaussian signal:

[0046]

[0047] 4) source signal is skew distribution signal:

[0048]

[0049] The kurtosis of the pipe network flow signal Q and the pressure signal P collected by sampling is calculated, and the super-Gaussian distribution is higher than 3, and the sub-Gaussian distribution is lower than 3;

[0050] (2) The observation data is initialized, and the steps are as follows:

[0051] 1) mean centering: calculate the mean of the samples x (1) , x (2) , … x (T) of the observation vector x, and subtract the mean for centering:

[0052] x-E{x}→x;

[0053] 2) whitening:

[0054] z=D -1 / 2 e T x;

[0055] Solve the eigenvalue and eigenvector of the covariance matrix E{xx T}, let e=(e1, e2, …e n ) be the unit normalized matrix composed of the eigenvectors as columns; D=diag(d1, d2…d n ) is a diagonal matrix composed of eigenvalues as diagonal elements.

[0056] Further, the constraint independent component analysis algorithm is combined with the constraint condition to perform iteration, including: determining the CICA iteration rule and the constraint condition, and the specific rules are as follows:

[0057]

[0058] Again, construct the Lagrange function, and use the Newton method to iterate to obtain:

[0059]

[0060] L' w =ρE{ZG y '(y)}-0.5μE{Zg y(w) = p E {Zy} - 0.5 mu E {Zg

[0061] (w) = p E {Zy} - 0.5 mu E {Zg yy (w) = p E {Zy} - 0.5 mu E {Zg y (w) = p E {Zy} - 0.5 mu E {Zg

[0062] (w) = p E {Zy} - 0.5 mu E {Zg k+1 (w) = p E {Zy} - 0.5 mu E {Zg k (w) = p E {Zy} - 0.5 mu E {Zg k (w) = p E {Zy} - 0.5 mu E {Zg k+1 (w) = p E {Zy} - 0.5 mu E {Zg k (w) = p E {Zy} - 0.5 mu E {Zg k (w) = p E {Zy} - 0.5 mu E {Zg

[0063] (w) = p E {Zy} - 0.5 mu E {Zg 2 (w) = p E {Zy} - 0.5 mu E {Zg k+1 (w) = p E {Zy} - 0.5 mu E {Zg

[0064] (w) = p E {Zy} - 0.5 mu E {Zg 2 (w) = p E {Zy} - 0.5 mu E {Zg k+1 (w) = p E {Zy} - 0.5 mu E {Zg

[0065] Another object of the present application is to provide a constant pressure water supply pipe network leakage rate evaluation system suitable for the constant pressure water supply pipe network leakage rate evaluation method, which comprises:

[0066] A blind source separation model construction module is configured to construct a blind source separation model of the daily water consumption and leakage of the water supply pipe network.

[0067] An observation data initialization module is configured to initialize the observation data.

[0068] A reference vector determination module is configured to determine reference vectors, including a water consumption signal reference vector and a leakage amount reference vector.

[0069] A water consumption signal and leakage signal acquisition module is configured to combine a constraint independent component analysis algorithm to iteratively obtain separated water consumption signals and leakage signals without actual physical meaning, with the water consumption signal reference vector and the leakage signal reference vector as constraint conditions.

[0070] An amplitude restoration module is configured to restore the separated water consumption signals and leakage signals to obtain water consumption signals and leakage signals with actual physical meaning.

[0071] In combination with all the technical solutions described above, the present application has the following advantages and positive effects:

[0072] The constant pressure water supply pipe network leakage rate evaluation method provided by the present application improves the robustness of the algorithm, thereby being applied to water consumption independent fields and sections such as industrial parks and stations to guide subsequent pipe network reconstruction plans.

[0073] The present application starts from improving the accuracy of the reference signal. The constant pressure water supply network pressure is stable, and the leakage of the leakage point is exponentially related to the pressure of the node, and the leakage is relatively stable, so a group of the same number of reference signals is constructed as the leakage reference signal.

[0074] Because the pipe network has only one water supply facility, the water consumption has little effect on the leakage, and under the above assumptions, the leakage signal is poor in Gaussianity, that is, the water consumption signal presents obvious Gaussianity, which also meets the premise of fast independent component analysis, and using the Lagrange proportional load method and Newton iteration method, two groups of signals can be quickly separated from the total meter flow signal and the pressure signal of the pressurizing equipment, and the one with high correlation degree with the flow signal is used as the water consumption reference signal.

[0075] Even if the current industrial parks, stations and other fields do not have automatic collection of flow and pressure remote devices, only need to install a common pressure gauge at the water pump during maintenance time, install a common water meter at the water pump outlet pipe, and send personnel to record the reading at regular intervals, the cost of collecting observation data is extremely low.

[0076] After obtaining the total meter flow and pressure signals for 24 hours, the water supply pipe network leakage evaluation work can be carried out.

[0077] As the creative evidence of the claims of the present application, it is also embodied in the following several important aspects:

[0078] (1) The expected income and commercial value of the technical scheme of the present application after transformation are: a leakage rate evaluation instrument suitable for constant pressure water supply pipe network can be formed, which is suitable for constant pressure water supply system, including but not limited to water tower water supply, constant pressure variable frequency water supply, pressure stable municipal pipe network water supply and other forms, which is convenient for factories, enterprises, parks, community properties to master their own pipe network leakage situation, calculate economic accounts, and make decisions on pipe network replacement plan.

[0079] (2) The technical scheme of the present application fills the domestic and foreign industry technical blank: there is no similar product in the current domestic and foreign market that can separate the water consumption signal and the leakage signal from the data of the inlet total water meter and the pressure gauge, the present application can be a new product, which is easy to install and apply.

[0080] (3) whether the technical solutions of the present application solve the technical problems that people have been eager to solve but have always failed to succeed: at present, many first- and second-tier cities such as Shenzhen, Guangzhou and Foshan and some railway stations have very serious pipe network leakage rates, and are carrying out or planning to carry out pipe network reconstruction projects, but how to reconstruct and from which places to reconstruct have always been unable to make macro decisions, and can only be artificially judged and cut off according to the service life of the pipe network, which causes a large amount of waste of funds, and some pipe networks that are relatively old in use but run well are blindly replaced, and some pipe networks that are not long in installation time but have serious leakage cannot be replaced and continue to maintain a high water leakage volume. BRIEF DESCRIPTION OF DRAWINGS

[0081] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0082] Figure 1 is a flow chart of a constant pressure water supply pipe network leakage rate evaluation method provided by the embodiment 1 of the present application;

[0083] Figure 2 is a flow chart of a constant pressure water supply pipe network leakage rate evaluation method provided by the embodiment 2 of the present application;

[0084] Figure 3 is a schematic diagram of a constant pressure water supply pipe network leakage rate evaluation system provided by the embodiment 3 of the present application;

[0085] Figure 4 is an inlet pressure signal graph provided in the experimental process of the present application;

[0086] Figure 5 is an inlet flow signal graph provided in the experimental process of the present application;

[0087] Figure 6 is a graph of separating the leakage amount trend signal and the real leakage amount signal provided in the experimental process of the present application;

[0088] Figure 7 is a graph of separating the water consumption signal and the real water consumption signal provided in the experimental process of the present application;

[0089] Figure 8 is a 24-condition comparison graph of the leakage amount provided in the experimental process of the present application;

[0090] Figure 9 is a 24-condition comparison graph of the water consumption provided in the experimental process of the present application;

[0091] In the figure: 1, blind source separation model construction module; 2, observation data initialization module; 3, reference vector determination module; 4, water consumption signal and leakage amount signal acquisition module; 5, amplitude restoration module. DETAILED DESCRIPTION

[0092] In order to make the above objectives, characteristics and advantages of the present application more apparent, specific embodiments of the present application are described in detail below with reference to the accompanying drawings. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced in many different ways from what is described herein, and skilled artisans will be able to make similar modifications to the specific embodiments disclosed herein without departing from the scope of the present application, and therefore the present application is not limited to the specific implementations disclosed below.

[0093] I. Explanation of Embodiments

[0094] Embodiment 1

[0095] As shown in the drawings, the present embodiment provides a method for evaluating leakage rate of constant pressure water supply network, which comprises the following steps: Figure 1

[0096] S101, using Lagrange proportional load method and Newton iteration method, separating water consumption signal reference vector and leakage signal reference vector from total table flow signal and pressurizing equipment pressure signal;

[0097] S102, referring to the reference vectors, respectively taking water consumption signal reference vector and leakage signal reference vector as constraint conditions, combining constraint independent component analysis algorithm for iteration, obtaining separated water consumption signal and leakage signal.

[0098] Embodiment 2

[0099] Based on the method for evaluating leakage rate of constant pressure water supply network provided in Embodiment 1 of the present application, as shown in the drawings, the method for evaluating leakage rate of constant pressure water supply network provided by the present embodiment further comprises the following steps: Figure 2

[0100] 1. Constructing blind source separation model of daily water consumption and leakage of water supply network:

[0101] x = As

[0102] x is an observation signal, which refers to constant pressure water supply equipment flow and pressure signal, which can be collected multiple times per hour, and the mean value after removing outliers is taken as the signal of the time period, one group of signals per hour, 24 groups per day.

[0103]

[0104] s is a source signal, which is the mean value of water consumption and leakage signal per hour, 24 groups per day.

[0105]

[0106] ​​Create a separation matrix W, such that y = Wx = WAs = Ks, when K tends to unit matrix, the separation signal y = s.

[0107]

[0108] According to central limit theorem, Σk i s i of Gaussian type is not weaker than any s i , the Gaussianity of y is weakest when and only when y is one of s. Obviously, in this case, k i in K is non-zero, y can be regarded as a successfully separated independent component. Gaussian distribution has the largest entropy value, construct the objective function to maximize the negative entropy of y.

[0109] Objective function: max J(y) = max (E{G(y)} - E{G(y guass )}) 2 = (E{G(w T z)} - E{G(y guass )}) 2

[0110] J(y) is the negative entropy of y; H(y) is the entropy of y, y guass is a Gaussian random vector with the same covariance matrix as y, with the maximum entropy. G(y) is generally as follows according to the degree of Gaussian of source signals:

[0111] 1) All source signals are super-Gaussian signals:

[0112] a is generally taken as 1.

[0113] 2) All source signals are sub-Gaussian signals:

[0114] G(y) = 0.25y 4 , G'(y) = y 3 ;

[0115] 3) Source signals are super-Gaussian signals and sub-Gaussian signals:

[0116]

[0117] 4) Source signals are skewed distribution signals:

[0118]

[0119] Calculate the kurtosis of the pipe network flow signal Q and the pressure signal P collected by sampling. Higher than 3 is super-Gaussian distribution, lower than 3 is sub-Gaussian distribution, the closer to 3, the stronger the Gaussianity, so as to select the appropriate G(y).

[0120] 2. Initialization of observation data, the steps are as follows:

[0121] 1) De-meaning: calculate the mean of the samples x (1) , x (2) , … x (T) of the observation vector x, and subtract the mean to center, that is:

[0122] x-E{x}→x;

[0123] 2) Whitening:

[0124] z=D -1 / 2 e T x;

[0125] Solve the eigenvalues and eigenvectors of the covariance matrix E{xx T}, let e=(e1, e2, … e n ) be the unit normalized matrix composed of eigenvectors as columns; D=diag(d1, d2…d n ) be the diagonal matrix composed of eigenvalues as diagonal elements.

[0126] 3. Determine the reference vector:

[0127] (1) The water quantity reference vector needs to construct a Lagrange function, and is obtained by iterative calculation using Newton method, and the following formula is obtained after simplification:

[0128] E{zG'(w T z)}+λw=0;

[0129] w i+1 =E{zG'(w i T z)}-E{G”(w i T z)}w i ;

[0130] w i+1 =w i+1 / ||w i+1 ||

[0131] If |w i+1 -w i |≤ξ, end iteration, output water signal reference vector r1=w i+1 z;

[0132] (2) The pipe network is a constant pressure water supply system, and the leakage quantity reference vector r2=ones(1, length(z)).

[0133] 4. Determine the CICA iteration rule and constraint condition, the specific rules are as follows:

[0134]

[0135] Reconstruct the Lagrange function, iterate using Newton's method, and simplify to obtain the following equation:

[0136]

[0137] L' w =ρE{ZG y '(y)}-0.5μE{Zg y '(w)}-λE{Zy};

[0138] δ(w)=ρE{ZG” yy (y)}-0.5μE{Zg' y (w)}-λ;

[0139] μ k+1 =max{0,μ k +γg(w k )};

[0140] λ k+1 =λ k +γh(w k )

[0141] When E{(yr) 2 If -ξ≤0, the iteration ends, and the output is y=w. k+1 z.

[0142] The algorithm described above is run twice, with reference vectors r1 for water usage and r2 for leakage, to obtain the separated water usage and leakage signals. Subsequent amplitude restoration steps can be performed using methods such as the minimum nighttime flow rate method.

[0143] Example 3

[0144] Based on the leakage rate assessment method for constant pressure water supply networks provided in Embodiment 1 of the present invention, such as... Figure 3 As shown, this embodiment of the invention also provides a system for assessing leakage rate in constant pressure water supply networks, comprising:

[0145] Blind source separation model construction module 1 is used to construct a blind source separation model for daily water consumption and leakage in the water supply network.

[0146] Observation data initialization module 2 is used to initialize the observation data;

[0147] Reference vector determination module 3 is used to determine reference vectors, including: water signal reference vector; leakage reference vector;

[0148] The water consumption signal and the leakage signal acquisition module 4 is used for referring to the vectors respectively to take the water consumption signal reference vector and the leakage signal reference vector as constraint conditions, combining the constraint independent component analysis algorithm to carry out iteration, and obtaining the separated water consumption signal and the leakage signal without actual physical meaning.

[0149] The amplitude reduction module 5 is used for reducing the separated water consumption signal and the leakage signal to obtain the water consumption signal and the leakage signal with actual physical meaning.

[0150] II. Application examples

[0151] At present, the pipe network leakage rate of many first- and second-tier cities such as Shenzhen, Guangzhou and Foshan and some railway stations is very serious, and pipe network reconstruction projects are being or planned to be carried out, but how to reconstruct and from which places to reconstruct cannot be macro-decided, and can only be artificially judged by the service life of the pipe network, which causes a large waste of funds, and some pipe networks that are relatively old in service life but are in good operation are blindly replaced, and some pipe networks that are not long in installation time but have serious leakage cannot be replaced and continue to maintain a high water leakage volume.

[0152] The current domestic and foreign markets do not have a similar product that can separate the water consumption signal and the leakage signal from the data of the inlet total water meter and the pressure gauge, and the present application can be a new product, which is simple to install and convenient to apply.

[0153] The technical scheme of the present application can form a leakage rate evaluation instrument suitable for constant pressure water supply pipe networks after transformation, which is suitable for constant pressure water supply systems, including but not limited to water tower water supply, constant pressure variable frequency water supply, pressure stable municipal pipe network water supply and the like, and is convenient for factories, enterprises, parks, community properties to master the pipe network leakage conditions of themselves, calculate economic accounts, and make decisions on pipe network replacement plans.

[0154] The information interaction, execution process and the like between the above devices / units can refer to the method embodiments based on the same concept, and the specific functions and the technical effects brought by the method embodiments can be referred to the method embodiments part, which will not be described here.

[0155] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or software. In addition, the specific name of each functional unit and module is only for easy distinction, and does not limit the protection scope of the present application. The specific working process of the unit and module in the above system can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

[0156] Application Example 1

[0157] The embodiment of the present application further provides a computer device, which comprises at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps in any of the method embodiments described above when executing the computer program.

[0158] Application Example 2

[0159] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, wherein the computer program is executable by a processor to implement the steps in any of the method embodiments described above.

[0160] Application Example 3

[0161] The embodiment of the present application further provides an information data processing terminal, which is used to provide a user input interface to implement the steps in any of the method embodiments described above when executed on an electronic device, and the information data processing terminal is not limited to a mobile phone, a computer, or a switch.

[0162] Application Example 4

[0163] The embodiment of the present application further provides a server, which is used to provide a user input interface to implement the steps in any of the method embodiments described above when executed on an electronic device.

[0164] Application Example 5

[0165] The embodiment of the present application provides a computer program product, which, when executed on an electronic device, enables the electronic device to implement the steps in any of the method embodiments described above.

[0166] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0167] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0168] III. Evidence of the relevant effects of the embodiments:

[0169] A certain project uses constant pressure water supply with a large leakage rate in the pipeline network. A water-saving and consumption-reducing plan was implemented, and water meters were installed at each water point to measure the actual water consumption curve (sum of water meter readings at each water point) and the actual leakage curve (total inlet meter - water meter readings at each water point) for 24 hours. These curves were used to verify the algorithm's separation results.

[0170] First, the inlet pressure and inlet flow signals were acquired from the source signal graph, such as... Figure 4 , Figure 5 Both are source signal graphs. After obtaining the reference vector, the algorithm is then used to separate the blind sources of water usage and leakage signals, resulting in a physically meaningless separated signal, such as... Figure 6 , Figure 7 Finally, amplitude restoration is performed to obtain a comparison between the actual leakage signal and the separated leakage signal, and a comparison between the actual water usage signal and the separated water usage signal, as shown below. Figure 8 Figure 9 .

[0171] The similarity coefficient between the water consumption separation signal and the source signal is 99.96%; the similarity coefficient for leakage is 99.92%.

[0172] Separation matrix Mixed matrix

[0173] As can be seen, since this example uses municipal constant pressure water supply, the leakage is basically a fixed value. The separation effect of the reference signal we selected is very good. Amplitude restoration and error analysis:

[0174] Since the signal y obtained from blind source separation has no real physical meaning, the amplitude of the separated signal must be restored if the true leakage amount is to be obtained.

[0175] For the data obtained from the example pipeline network, 24 operating conditions within a day were selected to perform amplitude restoration and obtain all the restored data values.

[0176] Since the separated signal is a scaled and translated version of the real source signal, the real value can be obtained by restoring the separated signal according to these steps.

[0177] First, separate the signal. (Where y1 is the separation leakage and y2 is the separation water consumption) are processed as follows to transform them into signals with a mean of 0 and a variance of 1:

[0178]

[0179] In the formula, l0(t) is a leak separation signal with a mean of 0 and a unit variance of 1.

[0180] y0(t) — A water separation signal with a mean of 0 and a unit variance of 1;

[0181] std[] — Calculates the standard deviation.

[0182] Therefore, the true separation value differs from l0(t) and y0(t) only by the mean shift and the standard deviation scaling transformation, i.e., there exists a relationship as shown in (2):

[0183]

[0184] In the formula Q L (t)——The leakage separation signal with the actual true value;

[0185] Q Y (t)——Water separation signal with actual true value;

[0186] σ L σ Y —The standard deviation of the actual separated signal;

[0187] μ L μ Y —The mean of the actual separated signal.

[0188] Adding the two equations in (2) together, we get equation (3):

[0189]

[0190] In the formula, Q(t) represents the inlet flow rate.

[0191] —The average inflow rate within sequence t.

[0192] Expand the time series t in (3) as shown in (4):

[0193]

[0194] Thus, (4) becomes solving for σ. L σ Y The overdetermined system of equations with two variables contains the constraints shown in equation (5):

[0195]

[0196] The genetic algorithm is used to solve the overdetermined system of equations. The fitness and objective function are defined according to equation (6):

[0197]

[0198] The population consists of 1000 groups, with 5000 iterations. The crossover and mutation coefficients are set to the default values ​​of the built-in ga function in MATLAB.

[0199] σ L =-6.4126, σ Y =0.0188.

[0200] At this point, we only need to find a set of known values ​​for water consumption and leakage in equation (2), and substitute them into equation (7) to obtain μ. L μ Y :

[0201]

[0202] In practice, the minimum flow rate at night is generally used for approximate estimation because there is almost no water usage for the project around 3 AM. After excluding factory water usage, the inlet flow rate at this time is very close to the leakage rate. In the case study, the leakage and water usage at 3 AM are used in the calculation, as the project has no water usage at this time, and the inlet flow rate is entirely leakage. After solving the overdetermined equations and substituting the leakage and water usage information at 3 AM, four unknown parameters σ are obtained. L σ Y μ L μ Y As shown in Table 1:

[0203] Table 1. Solution values ​​of the overdetermined equations

[0204]

[0205] Substituting the four unknowns into the overdetermined equation, we can obtain Q at each time step. L (t) and Q Y (t).

[0206] according to Figures 8-9 The results show that the average value of the water consumption amplitude is 11.7948, the actual average value is 11.7871, and the error is 0.065%; the average value of the leakage amplitude is 4.2995, the actual average value is 4.2987, and the error is 0.0186%. The separation effect is excellent.

[0207] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention and within the spirit and principles of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for assessing leakage rate in constant pressure water supply networks, characterized in that, The method for assessing leakage rate in constant pressure water supply networks includes: based on the characteristics of constant pressure water supply systems, using the Lagrange proportional load method and Newton's iteration method, separating water consumption signal reference vectors and leakage signal reference vectors from the total flow rate signal and the pressure signal of the pressurizing equipment, using these as constraints, and combining the constrained independent component analysis algorithm for iteration to obtain the separated water consumption signal and leakage signal. The method for separating the water usage signal reference vector from the total flow rate signal using the Lagrange proportional load method and Newton's iteration method includes: Construct the Lagrangian function E{zG'(w T z)}+λw=0; Using Newton's method iteratively, the iteration rule is as follows: w i+1 =E{zG'(w i T z)}-E{G”(w i T z)}w i ; In i+1 =in i+1 / ||in i+1 ||; After iteration, determine: |w i+1 -w i Does |≤ξ hold true? The step of iterating using the water signal reference vector and the leakage signal reference vector as constraints, respectively, combined with the constrained independent component analysis (CICA) algorithm, includes: determining the CICA iteration rules and constraints, the specific rules of which are as follows: Reconstruct the Lagrange function and iterate using Newton's method to obtain the following equation: L' w =ρE{ZG y '(y)}-0.5μE{Zg y '(w)}-λE{Zy}; δ(w)=ρE{ZG” yy (y)}-0.5μE{Zg' y (w)}-λ; m k+1 =max{0,μ k +γg(w k )}; l k+1 =λ k +γh(w k ) When E{(yr) 2 If -ξ≤0, the iteration ends, and the output is y=w. k+1 z.

2. The method for assessing leakage rate in constant pressure water supply networks according to claim 1, characterized in that, If |w i+1 -w i When |≤ξ holds, the iteration ends, and the water signal reference vector r1=w is output. i+1 z.

3. The method for assessing leakage rate in constant pressure water supply networks according to claim 1, characterized in that, If |w i+1 -w i If |≤ξ is not true, we can use Newton's method to iterate again.

4. The method for assessing leakage rate in constant pressure water supply networks according to claim 1, characterized in that, The leakage signal reference vector separated from the pressure signal of the pressurizing equipment includes: the pipeline network is a constant pressure water supply system, the leakage reference vector is a constant, and it is directly determined according to the constant pressure water supply. The leakage amount reference vector is r2 = ones(1, length(z)).

5. The method for assessing leakage rate in constant pressure water supply networks according to claim 1, characterized in that, Before separating the water usage signal reference vector and leakage signal reference vector from the total flow rate signal and the pressure signal from the pressurizing equipment using the Lagrange proportional load method and Newton's iteration method, the following steps are required: Construct a blind source separation model for daily water consumption and leakage in water supply networks: x = As x represents the observed signal, referring to the flow and pressure signals of the constant pressure water supply equipment. The signals are collected multiple times per hour, and the average value is taken after removing outliers as the signal for that time period. There is one set of signals per hour, for a total of 24 sets per day. s represents the source signal, the average hourly water consumption and leakage signal, with 24 sets throughout the day; Create a separation matrix W such that y = Wx = WAs = Ks. When K approaches the identity matrix, the separation signal y = s is achieved. According to the central limit theorem, Σk i s i Gaussian type ratio of any s i None of them are weak; the Gaussian property of y is weakest if and only if y is one of the elements in s; only k in K has this property. i If the value is non-zero, then y is considered a successfully separated independent component; the Gaussian distribution has the largest entropy value, so we construct an objective function that maximizes the negative entropy of y; Objective function: maxJ(y) = max(E{G(y)} - E{G(y)} guass )}) 2 =(E{G(w T z)}-E{G(y guass )}) 2 ; J(y) is the negative entropy of y; H(y) is the entropy of y, y guass G(y) is a Gaussian random vector with the same covariance matrix as y and has maximum entropy; G(y) has different Gaussian signals depending on the degree of Gaussianity of the source signal.

6. The method for assessing leakage rate in constant pressure water supply networks according to claim 5, characterized in that, The G(y) has different Gaussian signals depending on the degree of Gaussianity of the source signal, including: 1) The source signals are all super-Gaussian signals: 2) The source signals are all sub-Gaussian signals: G(y)=0.25y 4 ,G'(y)=y 3 ; 3) The source signals are super-Gaussian and sub-Gaussian signals: 4) The source signal is a skewed distribution signal: Calculate the kurtosis of the sampled pipeline flow signal Q and pressure signal P. Kurtosis greater than 3 indicates a super-Gaussian distribution, while kurtosis less than 3 indicates a sub-Gaussian distribution.

7. The method for assessing leakage rate in constant pressure water supply networks according to claim 1, characterized in that, Before separating the water usage signal reference vector and leakage signal reference vector from the total flow rate signal and the pressure signal of the pressurizing equipment using the Lagrange proportional load method and Newton's iteration method, it is necessary to initialize the observation data. The steps are as follows: 1) Mean-neutralization: Calculate the sample x of the observed vector x. (1) x (2) , ...x (T) Center the result by subtracting the mean from the mean: xE{x}→x; 2) Albinism: z=D -1 / 2 e T x; For the covariance matrix E{xx T To find the eigenvalues ​​and eigenvectors, let e = (e1, e2, ..., e2). n D is a unit-norm matrix composed of columns of eigenvectors; D = diag(d1, d2, ..., dn) n ) is a diagonal matrix whose eigenvalues ​​are diagonal elements.

8. A leakage rate assessment system for constant pressure water supply networks, utilizing the leakage rate assessment method for constant pressure water supply networks as described in any one of claims 1 to 7, characterized in that, The leakage rate assessment system for constant pressure water supply networks includes: Blind source separation model construction module (1), used to construct a blind source separation model of daily water consumption and leakage in water supply network: The observation data initialization module (2) is used to initialize the observation data. The reference vector determination module (3) is used to determine the reference vector, including: the water signal reference vector; and the leakage reference vector. The water consumption signal and leakage signal acquisition module (4) is used to obtain the separated water consumption signal and leakage signal without actual physical meaning by using the water consumption signal reference vector and the leakage signal reference vector as constraints respectively, combined with the constraint independent component analysis algorithm. The amplitude restoration module (5) is used to restore the separated water consumption signal and leakage signal to obtain water consumption signal and leakage signal with actual physical meaning.

Images